Professional Documents
Culture Documents
Jill Dych
Publisher: Addison Wesley
Acknowledgments
A business writer, trying to consolidate and make sense of the varied disciplines that comprise the intricate apparatus of even the simplest new technology trend, incurs many debts ! am grateful to those who have contributed to this book and proud to acknowledge their help "hanks first and foremost to the sub#ects of my case studies, and to those who guided me to them: $eth %eonard of &eri'on (ommunications) *onica "yson of +arrah,s -ntertainment) +arry -gler of -ddie $auer) .ari /pdal of 0nion $ank of 1orway) 2egina Wallace of Juniper $ank) and *ike /verly and -d *agin from +ewlett Packard "hese people had the generosity and authority to say yes "hanks too go to $ill Prentice and 1elle 3chant' of 3A3) .im 3tanick and (arrie $allinger from "eradata) 3usan Poser and 3usan 0mphrey from /racle) and *ary (han, fellow management consultant and world traveler *y reviewers represent the best and the brightest in the (2* industry, and share credit for this book,s real4 world perspective +ats off to $rian $erliner, Joy $lake, James (raig, (apers Jones, .elly *ooney, 2obin 1eidorf, *arcia 2obinson, David %inthicum, and Anne "homas *anes "hanks, too, to .im *outsos, editor of DB2 Magazine, for letting me reclaim te5t from an article ! wrote on clickstream analysis And to the vendors who provided e5amples of real, working products6$rian +oover of "ouchscape, 7adra 1ally and *ary 8ros of "eradata, 9ancy /shita from /racle, and John 8ill from (hannel Wave6you,ve enriched the +andbook immensely ! feel as if ! have a personal troop of (2* gurus in 7ran 7ra'er, %inda *c+ugh, and John -arle, who together corner the market on personali'ation, voice recognition and wireless technology, and (2* planning e5pertise, and who plowed through my drafts with the patience and understanding of good friends -van %evy lived through the writing of this book, revealing his own (2* battle scars at aberrant hours and without complaint !,m grateful to him for many things, foremost of which is that he,s still talking to me now that the book,s finished Addison4Wesley,s *ary /,$rien, whose imagination and creativity were the seeds of the +andbook, was indispensable in its development, as was editorial assistant Alicia (arey, copy editor e5traordinaire *alinda *c(ain, and the inimitable Addison4Wesley production staff And a final shout out to my colleagues at $aseline (onsulting 8roup, who picked up the slack when the going got tough, in particular 8ordon %evy and David 2ankell 9ou are (2* best4practices personified:
Introduction
/n one of those preternaturally warm spring afternoons, when many of their colleagues had forsaken them for the beach, around @>> conference attendees packed themselves into a hall at the %os Angeles (onvention (enter to hear about (ustomer 2elationship *anagement A group of high4profile e5perts was assembling to deliver a heralded panel discussion on the current and future state of the (2* market Attendance swelled to standing room only /n the panel were e5ecutives from both established and emerging (2* vendors /ne panelist headed a company that sold an !nternet storefront product Another ran a sales4force automation company A third represented a ma#or database vendor "here was a call center system vice president and, to his left, a chief privacy officer At the end of the line sat a renegade technology analyst
Chapter 1: +ello, !ntroduces (2*,s value proposition from a Any reader needing an introduction to (2* business perspective and e5plains why and its role in business strategy should read 8oodbye "he companies are rushing to #ump on the (2* this chapter 1ew 3pin on (ustomer %oyalty bandwagon
Chapter 3: (2* (overs why customer service is the locus of (ustomer support staff members at all levels most (2* programs and how new will en#oy comparing their company contact and (ustomer customer service strategies and center environments with the best practices 3ervice technologies promise to enhance customer outlined in this chapter Also of interest to loyalty6not to mention a company,s marketing staff considering other customer revenues touchpoints Chapter 4: 3ales "he birthplace of (2*, 37A includes a 7orce Automation variety of tactical and strategic functions "his chapter goes from managing customer leads and accounts to sharing customer knowledge via wireless media Chapter 5: (2* 8iven the challenges e4business presents, this chapter discusses where the customer in e4$usiness fits in the supply chain for both $=$ and $=( relationships Chapter 6: Analytical (2* Analytical (2* leverages the data gathered from cross4functional customer touchpoints to help companies make strategic decisions "his chapter covers the risks and rewards of analy'ing and acting on new customer knowledge 3ales managers and sales reps alike can use this chapter as a benchmark for how they,re managing their customer contacts and leads Also valuable for field service personnel 7or managers and developers responsible for delivering e4business, particularly e(2*, as well as users and developers of -2P and supply chain management systems 7or business people for whom decision support is a critical #ob function, as well as data analysts using sophisticated predictive techniAues Also helpful for marketing managers who rely on data analysis for launching new programs
Part II: Deli!ering CRM Chapter Chapter 7: Planning 9our (2* Program
Part II describes the key components of a (2* program and offers e5amples and checklists for ensuring they are performed thoroughly and in the right seAuence to mitigate risk and ensure successful (2* delivery Description -5plains how to evaluate your company against (2* critical success factors "his chapter also describes how to gauge the comple5ity of your (2* initiative and how that comple5ity determines a range of planning and development activities, including reAuirements gathering and 2/! calculation Discusses (2* technology software features and e5plains reAuirements4driven technology selection "his chapter contains checklists and interview Auestions for both (2* software vendors and application services providers <A3Ps? Intended Audience 7or business analysts and consultants who will be gathering and documenting (2* reAuirements, as well as pro#ect managers who will be charged with translating them into a working (2* system Also helpful for (2* sponsors and end users who must understand the tasks and resources necessary in (2* planning 7or !" e5ecutives and pro#ect managers charged with leading (2* technology selection efforts, as well as stakeholders who need to understand (2* technology4 selection best practices "he vendor evaluation Auestions might help vendors better prepare for prospect and client presentations "echnical staff, (2* development team members, and pro#ect managers will be interested in the roles integral to (2* pro#ects, as will (2* stakeholders who want to learn more about where to begin
Describes how to delineate, prioriti'e, and staff (2* pro#ects and highlights some common roadblocks to successful development Discusses establishing success metrics and measuring against them, and includes a (2* !mplementation 2oadmap
"his chapter introduces some of the main $usiness sponsors and pro#ect managers Chapter 10: interested in ensuring the success of their 9our (2* 7uture roadblocks known to sabotage (2* programs !t also covers some controversial (2* programs, as well as business users (2* trends who want a preview of (2* features on the hori'on
"oward the end of the content chapters, you,ll find a B(hecklist for 3uccess,C describing the best practices involved in achieving the ob#ectives discussed in that chapter <!f you,re underway with (2*, use this checklist as a tool to perform gap analysis against your current pro#ect ? !n addition, because (2* is inherently a business management initiative, each chapter concludes with a section titled B"he *anager,s $ottom %ine,C summari'ing the discussion for managers and e5ecutives who might be sponsoring (2* in their companies
%art I& 'e"ining CR# Chapter 1. Hello( )oodb!e& The *ew +pin on Customer ,o!alt!
"he so4called typical customer no longer e5ists, and companies have been learning this lesson the hard way 0ntil very recently, business was more concerned about the Bwhat,sC than about the Bwho,s C !n other words, companies were focused on selling as many products and services as possible, without regard to who was buying them *ost corporations cling to this product4centric view even today, basing their organi'ational structures and compensation plans on the products they sell, not the customers who buy "he $urger .ing slogan of the FGJ>s inviting the fast4food consumer to B+ave it your wayC was positively unorthodo5 for its era, as companies across industries offered standard products to the consumer population at large *oreover, FGK>s and FGJ>s corporate America bet their profits on classic marketing tactics6primarily television ads, mass mailings, and billboards6and then sat back and waited for the customers to pour in $ut the baby boomers came of age and competition burgeoned (onsumers had more choices than ever before about where to do their banking, their grocery shopping, and their vacationing Deregulation increased competition even further as it drove prices down (ompanies were forced to invent new methods of interacting with customers to reduce costs and gain market share 0se of automated teller machines <A"*s? and interactive voice response <!&2? systems increased $ut customers weren,t necessarily more satisfied than before -5ecutives soon reali'ed cost4reduction tactics weren,t enough to satisfy either customers or shareholders, who continued to call them on the carpet for eroding margins *a5imi'ing profitability was the real name of the game "he parado5 was that companies couldn,t very well increase profits while simultaneously enticing new customers with price breaks A"L" and *(! learned this the hard way in the FGE>s as they and other long4distance companies mailed millions of DF>> checks out to consumers, who switched their long4distance service and switched it back again
[1]
Although si5 times is acknowledged as the prevalent figure, reports on new4customer acAuisition costs vary, from as low as three times to as high as thirteen times
"he infrastructure that enables the delineation of and increase in customer value, and the correct means by which to motivate valuable customers to remain loyal6indeed, to buy again As we,ll see throughout this book, (2* is about more than simply managing customers and monitoring their behaviors (2* has the potential to change a customer,s relationship with a company and increase revenues in the bargain "he most forward4thinking companies have recogni'ed from past failures that (2* smacks of strategy, and thus technology alone can,t address high4profile issues such as new4customer acAuisition and Web4based marketing "o these companies, (2* is much more than a standalone pro#ect accounted for by a single organi'ation, it,s a !u"ine"" %hilo"o%h& that affects the company4at4large <We,ll see e5amples later of companies who practice (2* without even using the term ? "hese firms have articulated their ultimate visions for (2* to communicate them to every facet of operations "he following list represents a set of legitimate (2* business ob#ectives from several of my clients currently in the throes of their (2* programs: BWe want to thoroughly understand our customers, needs6even before they know them themselves C6A mid4market financial institution BDecreasing customer churn by increasing customer satisfaction C6A competitive local e5change carrier B*otivating customers to initiate revenue4generating contacts with us C6An online insurance company B!ncreasing the likelihood of the ,right response, by a given customer or customer segment C6A catalog retailer B"o use technology to improve customer service and enable a greater degree of customer differentiation in order to deliver uniAue customer interactions C6A data services firm BWe want to attract customers6both old and new6through more personali'ed communications C6 An online retailer
"he point here is that there is not one but many visions for (2* success (2* promises to help companies get to know their customers well enough to understand which ones to keep and which ones they should be willing to lose6and why6and how not to overspend in the meantime (2* also means automating many of the business processes and accompanying analysis and saving precious time in the bargain And saving money (harles 3chwab,s multimillion4dollar investment in 3iebel,s (2* product, which the brokerage firm uses to track each interaction with a customer or prospect, was recouped in less than two years [2] 3tories of wildly successful (2* programs have invaded both print and cyberspace, spurring otherwise cynical e5ecutives to turn their heads in the (2* direction After all, who could argueM
[2]
2oberts4Witt, 3arah % , B!t,s the (ustomer, 3tupid:C ' Magazine, June =J, =>>>, =K
Who indeedM "he estimated J> percent of companies who have tried implementing standalone (2* systems and failed, as well as the legions of other companies who have taken missteps on their e4business #ourneys might have a few comments As with enterprise resource planning <-2P?, supply chain management <3(*?, and other wide4reaching corporate programs that mandate a combination of innovative technologies, new business processes, and organi'ational buy4in, (2*,s failures are vast and visible
3weat, Jeff, B%ots of (ompanies Are "hinking About (ustomer 2elationship *anagement, $ut Progress (an $e &ery 3low6(2* 0nder 3crutiny,C Information Week via (/*"-N online, 3eptember F@, =>>>
!t certainly doesn,t take much for a consumer to turn her head to a competing product or vendor A #a'' buff has a mental list of the (Ds she wants to buy When (D 1ow e4mails her a discount code for the new Dave $rubek recording, she goes to the site and buys it despite her hefty Bwish listC on Ama'on com $ut #ust as loyalty is becoming the mantra on every e5ecutive,s lips, customer satisfaction rates are plummeting !t,s practically routine these days for consumers to vow never to do business with a particular merchant 2egardless of their frenetic embrace of the customer, companies seem to be angering customers at a faster pace than they are serving them !n June =>>>, (ortune columnist 3tewart Alsop wrote a scathing piece on 3print P(3 and its poor service "he column, titled BDear 3print: 9ou "icked /ff the Wrong 8uy,C provoked hallelu#ahs from 3print P(3 customers, one of whom responded: ! hate 3print and spend way too much time fantasi'ing about its demise ! have friends who have 3print too, and we talk to one another like members of a support group Whenever !,m in line at 3print stores, ! feel it is my duty to reach out to and dissuade as many prospective customers as ! can Another reader weighed in with this: "o list all my horrifying e5periences would render this letter too long for publication And another <with graphic metaphor?: !,m sure you,re getting a million thank4you letters from the rest of us who have been lied to, hung up on, over4billed, underserviced, and treated like cattle on the way to the slaughterhouse "reating customers like cattle is the antithesis of (2*, the goal of which is to recognize an) treat each cu"tomer a" an in)ivi)ual* "hat said, if one individual is dissatisfied, odds are he,ll tell a collection of other individuals6one widely accepted marketing rule4of4thumb claims the average unhappy customer tells eight other potential customers about his negative e5perience 3uch spreading of consumer disapproval turns the world of viral marketing, which depends on word4of4mouth from true believers, upside down <+iral marketing6a phenomenon in which consumer bu'' trumps advertising as the means of a product,s adoption6accounts for the popularity of such products as 2a'or scooters, The Blair Witch 'ro,ect, and *A( 3pice lip liner, to name a few ? Web sites such as www planetfeedback com and www downside com are e5panding the reach of these Bviral complainersC and even speculating on the demise of companies that proffer poor service "he influence of such groups could in fact impact whether a product, indeed an entire company, succeeds or fails
+cenario
9ou,ve spent two grueling weeks of nonstop business in %ondon and are ready to head home &irgin Atlantic Airways sends a driver to fetch you at your hotel and bring you to the airport 0pon arrival at +eathrow, the driver stops at an outdoor kiosk 9our window magically rolls down to reveal a uniformed &irgin associate, who politely reAuests your ticket As the associate checks you in, the driver retrieves your luggage from the trunk6the Bboot,C in the local vernacular6tags it, and deposits it on the baggage conveyor belt "he &irgin associate smiles and hands you your boarding pass
!n FGGG, JJ percent of customer Auestions were handled online and customer satisfaction levels increased by => percent over FGG@, according to u"tomer /ervice on the Internet by Jim 3terne <Wiley, =>>>?
!nternet users appreciate not having to go out of their way to buy what they want, and the simpler the process, the higher the potential for customer satisfaction !ndeed, the Web offered customers options they hadn,t had with other delivery channels, namely: =H4hour access 0p4to4the4minute information <on, for e5ample, stock levels, product features, and prices? "he ability to research a product or merchant )uring a shopping trip /nline customer support /nline self4service Personali'ed content
(onsider the old way A customer needs a new set of window blinds +e goes to the kitchen, finds the yellow pages, and calls his local blind company "he representative e5plains they can have someone come out and measure for the blinds a week from "hursday "he rep arrives from the blind company, measures, and shows the customer photos of various blind styles and colors "hen the rep takes another two days to write up an estimate Almost two weeks later the customer has the pricing information he needs 1ow he must decide whether to get another estimate or take his chances with the only vendor he,s contacted (ompare that to the Web version "he customer enters Bmini blindsC in his favorite search engine, which returns the Web sites for several catalog window treatment firms +e chooses a company, which displays a series of blind designs and prices4per4inch "he customer chooses a design he likes, enters his window measurements, and receives a price online, including ta5 and shipping $efore purchasing, he browses a couple of other window4ware Web sites for additional blind designs and prices, eventually placing his order in less than an hour Although basic, the above e5amples illustrate why the Web has made doing business easier than ever 7or the customer, the time savings6even if he has to measure the window himself6is improved by orders of magnitude Plus, while he awaits delivery he can revisit the Web site to validate measurements or refresh his memory on his chosen style 7or the blind retailer, the order arrives electronically, decreasing sales time while reducing the chance of error <"he measurements, after all, are now the customer,s responsibility ? 7urthermore, the company has captured other information6the customer,s preferred blind style, his neighborhood, his interior color scheme 6for potential use in follow4up communications At the =>>> (omde5 conference in %as &egas, (isco (-/ John (hambers put a new spin on leveraging customer loyalty with the !nternet (hambers demonstrated a Web4connected gas pump that allowed a customer to swipe a Bloyalty card C 3wiping the card not only starts the gas flowing, it also illuminates a digital screen displaying personali'ed messages such as traffic reports while the customer pumps her gas A customer,s initial reluctance to subscribe to such services can be more than offset by the value they provide
5hat's In a *ame6
$eing the )e rigueur bu''4term of the moment, (2* was bound to spawn similar acronyms 7ollowing are some (2*4related terms you will come across in this book and elsewhere in your (2* readings: e RM 0alternativel&1 e# RM2* e(2* refers to BelectronicC customer relationship management or, more simply, (2* that is Web4based 7or instance, when you log on to golf retailer chipshot com to see if they,ve shipped your new titanium driver, that,s e(2* E RM 3omewhat confusingly, many e5perts and publications are using -(2*, alternatively coined B-2*,C to refer to BenterpriseC (2*, meaning a (2* program that spans an enterprise4wide view of a customer <We,ll spell out enterprise (2* to avoid any confusion with electronic (2* ? 'RM BPartner relationship managementC allows a company to manage its alliance partner and reseller relationships to provide customers with the optimal sales channel while streamlining the
!rrespective of the type of (2* a company is looking into, the common denominator is motivating the right customers to continue doing business with you We,ll see the various types of (2* at work in the ne5t several chapters Another important (2* distinction is BoperationalC versus BanalyticalC (2* "he distinction is an important one, because it speaks to the tactics a company is taking in implementing its (2* strategy As we,ll see in Part II of the +andbook, there is no one right answer to (2* implementation as long as business reAuirements drive the initiative O%erational RM1 also known as Bfront4officeC (2*, involves the areas where direct customer contact occurs We,ll refer to these interactions as customer Btouchpoints C A touch%oint can be an inbound contact 6e g , a call to a company,s customer support hotline6or an outbound contact6e g , an in4person sales call or an e4mail promotion "he ma#ority of self4described (2* products on the market today fall into the operational category Figure 1-1 illustrates the various levels of operational (2* 0igure 1-1. 2perational CR#& Touching the customer
/perational (2* enables and streamlines communications to and from customers, but this doesn,t necessarily mean optimi'ing service Just because a banking customer checks her balance on your Web site won,t conclusively establish that she doesn,t prefer to perform her transactions in the branch $ut how do you know for sureM Anal&tical RM1 also known as Bback4officeC or BstrategicC (2*, involves understanding the customer activities that occurred in the front office Analytical (2* reAuires technology <to compile and process the mountains of customer data to facilitate analysis? and new business processes <to refine customer4facing practices to increase loyalty and profitability? 0nder pressure from analysts and industry e5perts, most of today,s (2* vendors are either creating analytical (2* capabilities or partnering with business intelligence <$!? vendors to incorporate analysis into their offerings Figure 1-2 shows how the data and processes combine to refine business actions
As we,ll see throughout the +andbook, the refined4business4actions piece of the pu''le is the most difficult of all to put in place "he greater the number of missing pieces, the harder it is to construct a meaningful (2* picture Put another way, if enhanced customer loyalty is the door, integration is the key
&aluable as they are, none of these capabilities reAuires a (2* product !n fact, companies from brokerage houses to pharmaceutical firms were performing these duties long before the (2* acronym came along "he combination of data warehouses and analytical toolsets has given companies the ability to drill down into integrated data to reveal interesting6even competitively differentiating6findings 2ather than e5trapolating what types of promotions to launch and guessing who would respond, companies have begun relying on business intelligence analysis to provide them with hard facts that help them make better, more informed decisions and reap unforeseen rewards
B2ules repositoryC refers to the storage of business rules that describe data and how it,s used 7or instance, a business rule that ensures a discount for a high4value customer could be e5ecuted at the time the customer places an order or could simply define that the term BrevenueC refers to booked revenue rather than billed revenue
Although often misrepresented, the differences between business intelligence and (2* are distinct 9es, they both involve critical business decisions and both rely on information technology to deliver value "he e5amples in Table 1-1 illustrate the distinction 9our first impression might be that (2* is more comple5 than business intelligence !n fact, at most companies the number of true (2* users is a mere subset of the business population using business intelligence +owever, business intelligence, when not e5ploited to its full potential, can result in analysis for analysis, sake "he ma#or difference between $! and (2* is that RM integrate" information with !u"ine"" action* !n each of our e5amples, the (2* action will be tested and further refined (2* combines data analysis with the deployment of specific business actions "he ability to access data is, by itself, immensely powerful, but many business intelligence environments simply use data to confirm already held hypotheses "he mandate of (2* is the ability to act on that data and to change fundamental business processes to become more customer4centric Table 1-1. Business Intelligence 9ersus CR# Business Intelligence CRM CRM Rationale Display the name and address Display "ech(o,s most recent inbound contact 3alespeople become aware of business customer "ech(o on my personal digital assistant <PDA?, along of e5isting or in progress with their current corporate address issues before meeting with the customer Display customers who visit one of the video stores in our chain on a weekly basis /nce a month for the ne5t si5 months, send a (onvert casual visitors to direct4mail solicitation to customers most likely freAuent visitors to rent ne5t month,s new features who are not weekly visitors to the store 7ocus on retaining high4 value customers
Display a list of customers who (ontact all high4value customers who have have lodged a complaint within lodged a complaint 8enerate retention the past I> days recommendations for each customer <using (2* product feature? Analy'e the top five most popular office supplies and com4pare approved vendors, prices to prices of other potential suppliers
!dentify the top five purchased office sup4plies !ncrease the likelihood of and trial4run an automated Web reAuest4for4 price improvements on Auote <27O? system for limited Auantities to test commodity purchases price improvements 2eward repeat customers who are profitable and gather valuable prospect behavior data
%ist the e4mail addresses for 3end profitable registered customers a D@ registered customers who online discount if they fill in a form e5plaining abandoned their shopping carts why they abandoned their shop4ping carts during their last Web visit 3end F> percent off to unknown visitors if they complete the form
Product marketing campaigns were designed to reach as many people as possible *ass marketing campaigns not only treated all customers as if they had the same needs and wants, but also focused on the product being pitched, not on its potential buyer *ass marketing a product means stocking the same shelves across all stores with the same product and alerting potential customers in the same way "he underlying assumption was that the availability and choice of products a company offered consumers was the principal determinant of its success A product focus nevertheless does not imply a lack of market understanding /n the contrary, as companies launched and re4launched their products, they began analy'ing the responses to determine the success of marketing campaigns $usiness #ournals wa5ed poetic on product distribution strategies, and statisticians were suddenly in demand as companies began institutionali'ing such data analysis practices as product revenue analysis and price4elasticity modeling "he FGK>s introduced the phenomenon of Bdirect marketing C Direct marketing involved communicating with large numbers of consumers6usually through promotional mailings or maga'ine advertisements6and inviting them to respond to the promotion <B3end check or money orderPC? by ordering the product through the mail $ased on the principles of mass marketing, direct marketing usually focused on selling a mass4 produced product, be it E4track tapes, porcelain plates, or glass4encased stamp collections, to as many consumers as possible $ut direct marketers were pioneers in one respect: they monitored responses to their ads, often creating several versions of the same campaign and launching it in different areas of the country "hey then analy'ed relative response rates, whittling away the marketing messages that didn,t work and refining other campaigns based on those learnings Direct marketing stood the world of conventional advertising on its head, for the first time suggesting that messages could be tailored and consumers had preferences Despite its sophistication relative to the more general mass marketing campaigns, a direct marketing promotion was considered a smash if it generated a = percent response rate !n fact, according to the Direct *arketing Association, if a company,s direct mail campaign touches F>>,>>> customers and costs DGE,>>>, the company nets a significant loss [1]
[1]
As reported in
Target #arketing
As consumers began purchasing and using products, more data became available about them Data analysts began associating products to the customers who were buying them And it was through these analysis activities6as well as rising temperatures in the competitive climate and a few timely maga'ine articles6that companies began to understand that their <usually inert? customer data could be as valuable as the product data they were busy probing "hus evolved the term Btarget marketing,C the practice of promoting a product or service to a subset of customers and prospects
!n *arch of FGKH, author Daniel 9ankelovich wrote about B1ew (riteria for *arket 3egmentationC in the 3arvar) Bu"ine"" Review1 arguing that demographic segmentation, while valuable, was not the only way a company could categori'e its customers
*odern companies assign a variety of segments to their customers, often dynamically defining segments and temporarily regrouping customers for specific campaigns *arketing departments routinely group customers into the following categories: 8eography or regionality Psychographics[3]
[3]
BPsychographicC segments define groups of customers with similar interests, opinions, and preferences
7irmographics[4]
[4]
B7irmographicsC characteri'e a business and are used especially often in business4to4business <$=$? communications
!nfographics[5]
[5]
B!nfographicC segments delineate customers according to how they want to be communicated with <via e4 mailM direct salesM a proprietary Web siteM?, as well as how they prefer to interact with the company
Preferred sales channel Profitability 1umber of products 3ales territory "enure %ifetime value +ousehold demographics 2isk score %ife stage Privacy preferences
$y segmenting customers, companies could begin more speciali'ed communications about their products *uch of this relies on the company,s understanding its business strategies to the e5tent that it knows its most desirable segments 7or instance, if a bank has set its sights on deriving most of its profits from fee4 income products offered in its investment services line, customers of this organi'ation will likely have different preferences and characteristics from those opening savings accounts 3egmenting customers based on their preferred line of business or desired product features can reveal interesting facts about their different preferences and behaviors 7or instance, $oots, a 0 . retailer, segments customers who respond to promotions based on whether they are Bdeal seekers,C BstockpilersC <who buy in bulk when an item goes on sale and then don,t visit the store until they need to stock up again?, loyal e5isting buyers <whose purchase patterns increase for sale items?, or new market customers <who buy a sale item and continue to buy it after it reverts to its regular price? [6]
[6]
"he advantage of target marketing is that, if the company has @> million customers, only FH million or so6 =E percent6will receive the promotional mailing for digital service At a bulk rate of eight cents per mailing, target marketing saves the company almost DI million in postage alone <"he company might incur labor and mail service costs, which would also be sharply reduced ? "his allows the cable company to steer clear of disinterested prospects and avoid the all too common practice of over4communicating to customers !t fosters the customer,s perception that marketing is a tailored service, not an annoyance "hrough the significant cost savings and increased response rates, target marketing #ustified increased e5penditures for information technology !n fact, the better companies became at analy'ing data about their customers, the more effective their campaigns became, creating the closed4loop effect illustrated in Figure 2-2 0igure 7-7. Re"ining marketing campaigns
Figure 2-2 illustrates that the more a company knows about its customers, the more detailed its marketing messages can be 2efined marketing campaigns, targeted to customer segments rather than to the mass populace, became the marketing +oly 8rail
Peppers, Don and *artha 2ogers, Ph D , The One to One (uture$ Buil)ing Relation"hi%" One u"tomer At a Time 1ew 9ork: Doubleday, FGGI
Peppers and 2ogers rounded out marketing,s evolution from mass4marketing standard products to segmenting customers to true relationship marketing, or one4to4one Figure 2-3 shows the stages and their differences 0igure 7-:. #arketing e9ol9ing
/ne4to4one means not only communicating with customers as individuals, but also developing custom products and tailored messages based on customers, unspoken needs !t relies on a two4way dialog between a company and its customers to foster a true relationship and allow customers to truly e5press the desires the company can help fulfill !t relies as heavily on the customer,s e5perience of the company as it does on the specific marketing messages he receives "he concept of Bpermission marketingC goes hand4in4hand with one4to4one: customers not only need to be communicated with as individuals, they themselves should be able to stipulate how and when After all, as 3eth 8odin says, B9ou can,t build a one4to4one relationship with a customer unless the customer e5plicitly agrees to the process C[8] Permission marketing implies that the customer will decide when and for what reason she will accept to be interrupted with a marketing message *ost of the time, it,s when the customer perceives there is value to her in such an interruption A current e5ample is Ama'on com,s notification of /prah,s new $ook (lub selection to subscribers who have opted in
[8]
8odin, 3eth, 'ermi""ion Marketing$ Turning /tranger" into (rien)"1 an) (rien)" into u"tomer" 1ew 9ork: 3imon and 3chuster, FGGG, K@
Campaign #anagement
!f traditional marketing is, as the saying goes, part art and part science, the marketing of the future is pure science with a bit of flourish "he success of corporate marketing programs has become directly proportional to a company,s ability to capture and analy'e the right data !n days of mass marketing with its Bbatch and blastC mentality, most companies would decide on a marketing campaign based on someone,s interesting idea A team of product managers would plan the campaign, coming up with hit4or4miss media guesses and sound bites "hen they,d pull the ripcord and pray !t could take months to figure out whether a campaign was working, and months more to repair and re4launch it Why the throw4it4at4the4wall4and4see4if4it4sticks approachM 7or one thing, finding, entering, storing, and tracking individual customer data records was beyond most companies, capabilities, not to mention their budgets Parado5ically, the larger the company, the greater the number of customers and thus the more cumbersome and e5pensive managing customer data became %arge companies began undergoing the +erculean task of collecting and storing customer data in primitive databases !n the FGE>s, the advent of relational databases and data warehouses <databases e5pressly designed to hold large amounts of corporate data? promised to ease the pain of amassing detailed customer records and availing them to business users Chapter 6 discusses how data warehouses have gone from a large4company lu5ury to a (2* necessity Another challenge was the sheer labor involved in launching new marketing campaigns, rendered all the more difficult with the increasing freAuency of smaller, more targeted promotions *arketing departments needed to conceive of and plan the campaign, define the targeted consumers, determine the channel6how the campaign would be communicated6and then launch the promotion, as shown in Figure 2-4 0igure 7-;. A linear campaign management process
$ut only those with the necessary data can really monitor the results of that campaign, and use those results to refine future campaigns, a process known as Bclosed loopC campaign management "his means the ability not only to analy'e and understand whether a campaign is a success and why !t means using that new knowledge as the basis for future campaigns, in effect closing the loop with evolving customer information, as in Figure 2-5 0igure 7-<. Closed-loop campaign management
!t could be months or even years before companies understood whether a marketing campaign was successful 1o wonder the practice of campaign management mandated large marketing organi'ations and even larger marketing budgets: *any companies relegated the cumbersome work to marketing service bureaus or direct marketing agencies, whose turnaround time was less than stellar *any stayed in their marketing comfort 'ones, e5ecuting ad4hoc, point4in4time campaigns that generated short4term revenues but did nothing to reveal customer preferences or improve customer relationships "he time and cost being spent on these campaigns and the opportunity cost associated with ad hoc or e5perimental promotions were noted by technology vendors, many of them database product companies, who began devising ways of automating key pieces of the campaign management process "he initial developments in campaign management software were list generators %everaging current customer information in e5isting databases, list4generation software dynamically segmented customers for certain campaigns, using standard database Aueries "he result was a list of customer names and addresses that fit the specified boundaries 7or e5ample, a marketing department at a bank wants to offer freAuent depositors with large balances access to low4interest lines of credit "he specification BDisplay the names and addresses of all customers with checking account balances of over D=@,>>> who have an average deposit amount of D@>>> or higherC could easily be translated into a Auery and submitted to the bank,s customer database "he resulting list would become the basis for a mailing offer to those customers "he bank saves money not only by applying the principles of target marketing and segmentation6communicating to a proscribed and thus smaller group of customers6but also by avoiding the cost and time involved in searching out those customers manually As campaign management software became more popular, its features became more robust "he richer a company,s customer database, the more it could leverage that data to actually understand customer behaviors and institute true one4to4one marketing Products such as Nchange and Prime 2esponse began addressing components of the full campaign lifecycle, including (ampaign definition Planning (ustomer segmentation 3cheduling 2esponse management
*ost campaign management products aid the user in defining a workflow, including a schedule for the campaign "his allows a marketing department to delineate discrete tasks, specify costs and e5pected returns, and test various scenarios on different customer segments, campaign schedules, and product offers 3uch products not only shortened the marketing lifecycle, they enabled companies to dedicate more time to customers and less on managing the marketing process 7reeing up time in turn allowed campaigns to be launched in parallel instead of serially 7or instance, a telephone company,s single, monolithic custom4calling campaign could become several more4targeted campaigns featuring voicemail, caller4id, call waiting, or packages combining these and other products to specific customer segments *ultiple simultaneous promotions resulted in an increase in overall response rates, often in the double digits $y increasing not only the Auality but also the volume of their marketing campaigns through marketing automation, companies were increasing their revenues (2* products offering campaign management present savings of time, labor, and cost and often demonstrate Auantifiable returns on investment, as they have to the following companies: An outdoor products catalog company that customi'ed catalogs for specific customer segments <segments include BDay4trippers,C B(lotheshorses,C and B2oughnecksC? has boosted its average purchase by segment to =I percent, thereby increasing its overall catalog revenues by a staggering @>> percent An online brokerage firm,s cross4selling of its new checking account offering to holders of .eogh retirement accounts generated a I= percent response rate from its targeted base of customers, versus H percent from a control group An e4tailer speciali'ing in lifestyle products saw sales of its languishing line of house4brand women,s clothing nearly double when the company promoted select casual outfits with a well4known brand of baby products favored by new mothers
When mastered, automated campaign management is like a good wine: it gets better with age With marketing campaigns targeting increasingly smaller customer segments, companies can increase the number of campaigns simply to ensure that they continue interacting with as many customers as possible And by tracking the success of campaigns over time, companies can refine them, resulting in even higher response rates and thus greater corresponding revenues "he practice of closed4loop campaign management6using the results of campaigns to refine future campaigns6has become an acknowledged (2* best practice
Customer Retention
!n FGGK, author 7rederick 2eichheld wrote that 0 3 corporations lose half their customers every five years [9] !ndeed, banks and telephone companies were already in the throes of analy'ing customer attrition to understand why customers were leaving for competitors
[9]
2eichheld, 7rederick 7 , B%earning from (ustomer Defections,C 3arvar) Bu"ine"" Review1 *arch4April FGGK
0nderstanding that customers have left, and knowing specifically who, is non4trivial 0nderstanding why they have left is even more difficult +arder yet is stemming the tide of customer attrition by applying this knowledge to business tactics that encourage customers to stay Analy'ing customer attrition operates on the aphorism established in Chapter 1 that keeping an e5isting customer is far more cost effective than acAuiring a new one After all, the more customers leave, the greater the loss of revenue, loss of the initial acAuisition investment, and loss of a stable market base for selling new products Although this business tenet hasn,t arrested the fren'y to acAuire new customers, it has resulted in millions of dollars being spent to analy'e not only why customers leave, but also which customers are the most desirable and thus worth keeping !n the FGE>s, companies began focusing on who had left !n the FGG>s, they began applying characteristics of customers who had left to e5isting customers, thereby pinpointing those customers who might be leaving ne5t 2educing customer defections by even a fraction has been proven to increase profits e5ponentially (urrently, companies are using sophisticated predictive technologies that compare like attributes of similar customers to delineate customers who are Blikely to churn,C and they,re simultaneously personali'ing tailored marketing interactions designed to motivate those customers to stay "he trouble with customer retention strategies is that, once you,ve identified customers who might leave, how do you keep themM Despite sophisticated churn prediction products, most companies still don,t really know whether product giveaways and add4on perks6whose initial cost outlay often e5ceed the resulting revenue contribution6really result in profitability Will the cost of the free mobile phone be recouped in the customer,s service spendingM +ow often are customers continuing their digital cable subscriptions after the first I <free? monthsM Will that first4class upgrade ensure that a given customer will fly with you againM While they figure out the best way to keep customers who are on the brink of leaving, companies are working on designing marketing campaigns to bump low4value customers to a higher value band rather than allowing them to churn by default (hurn prediction is one of several uses of technology and data that aid marketing in predicting customer behavior
Beha9ior %rediction
Although not so much a marketing practice as a marketing enabler, behavior prediction helps marketing departments determine what customers are likely to do in the future 0sing sophisticated modeling and data mining techniAues6we,ll discuss some of these in more detail in Chapter 66behavior prediction uses historical customer behavior to foresee future behaviors "his analysis includes several variations: 'ro%en"it&#to#!u& anal&"i"* 0nderstanding which products a particular customer is likely to purchase 5e-t "e6uential %urcha"e* Predicting what product or service a customer is likely to buy ne5t 'ro)uct affinit& anal&"i"* 0nderstanding which products will be purchased with other products Also known as Bmarket basket analysis,C it can be viewed as e5amining products in a shopper,s basket to understand possible product associations 'rice ela"ticit& mo)eling an) )&namic %ricing* Determining the optimal price for a given product, often for a given customer or customer segment
$y understanding how a customer is likely to behave, a company can make a host of marketing decisions based on this knowledge, including these: Preemptively offering discounts or fee waivers to e5isting customers who are at risk of churning 2efining target marketing campaigns to smaller customer segments or specific products
"he key to all this analysis, and especially to the actions that result, is knowing who your best customers are
Channel 2ptimi3ation
"he goal of marketing automation is to offer the right message to the right customer at the right time With the advent of the !nternet, many firms are appending Bthrough the right channel,C to this ma5im <denoted by BDetermine +owC in Figure 2-4? as customers, interaction preferences evolve 7or instance, a new customer whose use of online banking services has steadily increased might prefer to be e4mailed a new offer along with her regular statement, whereas a retiree who en#oys visiting the neighborhood branch might be delighted when the branch manager offers her a cup of coffee and a brochure on a new annuity product !ndeed, banking customers have a range of choices when it comes to their preferred channels, as Figure 2-6 illustrates 0igure 7->. =arious banking ser9ice channels
0nderstanding the channels through which specific customers prefer to interact with your company is only a slice of the pie 9our company must also decide how best to communicate with your customers Just because a valuable customer prefers making his deposits in the branch doesn,t mean he won,t demand comprehensive online statements (hannel management means optimi'ing a company,s BinboundC channels with its BoutboundC means of customer interaction and knowing how to choose the best approach for each
%ersonali3ation
Practically all of us have purchased something over the !nternet, be it books, large appliances, or a pi''a $ut have you ever noticed that sometimes one of those Web sites seems to be talking right to youM Personali'ation is the capability to customi'e customer communication based on knowledge preferences and behaviors at the time of interaction !t can take any of the following forms: 9ou log on to your favorite music Web site "he site greets you with B+i, Jill, welcome back: We have some new picks for you:C 9ou scroll down and see a variety of new (Ds you,ve had your eye on but have been too busy to browse through at the record store 9ou double4click on the cover of the new $eastie $oys anthology and listen to one of the cuts While channel4surfing, you come across a hedge trimmer being featured on a home shopping station 9ou know your wife would love it 9ou call to order it and the operator greets you by name and asks for the item number After confirming that the hedge trimmer can be sent overnight, she asks you whether you,d like to use your &isa card 3he already has the number 9ou say yes </bviously, your wife has shopped here before ?
Although both of these e5amples are leveraging personali'ation, the first e5ample is personali'ation as most people know it: online messages tailored to a particular customer or customer segment 3uch tailored messages can involve anything from inserting the Web shopper,s name into the message6 known as Bvariable insertionC6to using detailed customer data to personali'e Web site content 7or instance, a cosmetics e4tailer might want to promote de4fri''ing shampoo to shoppers in the humid southeast and a conditioner with 3P7 protection to shoppers in Palm 3prings *ore specifically, personali'ation technologies can tailor messages to individual customers, accessing current personal data each time the customer visits the site and using it to create custom content "hese technologies enable analysis of each customer over time and across all channels, using customer profile data, past purchases, clickstream data, and Web survey responses to determine, for instance, what product the customer is most likely to purchase ne5t or whether the customer is at4risk and thus deserving of a
Figure 2-8 shows the first screen as it would be seen by the prospect, having been Bfilled in,C with the actual offer "he prospect is proffered additional high4margin products, allowing the bank to ma5imi'e the opportunity for cross4selling 0igure 7-@. A personali3ed marketing message .Courtes! o" Teradata CR#/
Personali'ation technologies can apply their learnings6if the customer responded to a discount on skis, it,s likely she,ll be interested in cold4weather apparel6to future personali'ed messages, having proven improved response rates over time as the customer,s behavior data becomes ever more enriched "his removes the guesswork, resulting in a creeping understanding of customers and their preferences over that customer,s relationship with your company .nown as Badaptive personali'ation,C it,s a critical factor in making it as difficult as possible for your customer to switch to a competitor 7or e5ample, Wine com uses personali'ation to tailor uniAue e4mail newsletters to certain customers based on past purchases Personali'ation in the $=( space is largely based on the analysis of a customer,s clickstreams, his navigation path through a company,s Web site $y monitoring a customer,s clickstream, a company can see not only what a customer purchased but how the customer reached the site in the first place <a partner Web siteM A banner adM?, how he traveled through the site after he got there, how much time he spent on each page, and which products might have stimulated purchases of other products !n short, clickstreams can shed light on a customer,s Web e5perience 7or now, it,s important to note that analy'ing clickstreams can trigger do'ens of new tactics, including (hanges to Web impressions <images on the Web site? according to a customer,s navigation patterns and past purchases (ustom promotions or discounts based on past purchases or research (ustomi'ed Web pages according to the visitor,s use of the site
Chapter 6 features a real4world scenario of clickstream analysis at a specialty e4tailer and how it presented a choice of business actions
49ent-Based #arketing
"he best definition of event4based marketing is a time4sensitive marketing or sales communication reacting to a customer4specific event -vent4based marketing6also called event4driven marketing6can apply to a segment of customers or to individual customers 7or instance, mailing an application for an increase in collision damage insurance to all customers who,ve recently had traffic accidents is an e5ample of event4 based marketing to a largely undifferentiated segment of the e5isting customer base
B(ustomer Affinity (lub "argets 7irm,s %oyalists,C The Wall /treet Journal1 1ovember F@, =>>>
3herman, -rik, B"inker, "ailor, 3oftware, 3py,C 5ew"week1 /ctober FK, =>>>
And companies who blow the privacy issue are getting unwanted attention for often4innocent errors !t,s not only the ubiAuitous consumer protection agencies vying for more privacy restrictions, but government regulators as well "he 8ramm4%each4$liley Act in the 0nited 3tates established mandatory measures for 0 3 financial institutions and their use of data, and many consumer advocates believe it,s the harbinger of privacy legislation for other industries Almost all -uropean countries now have federal agencies dedicated to protecting privacy, with the -uropean -conomic 0nion guaranteeing individual compensation in instances of privacy abuse B(onsumers shouldn,t have to reveal their life story every time they surf the Web,C said (alifornia congresswoman Anna -shoo in a press release announcing her proposed legislation [12] reAuiring companies to adopt Bopt inC policies "he 3helby Act recently barred state motor vehicle agencies from sharing driver registration information without e5plicit consent, reversing the previous opt4out policy (orporations fear such legislation because it will reAuire them to receive e5plicit customer permission to solicit and use their personal data 1ot only will such opt4in policies thwart many e5tensive and intricately planned marketing strategies, they will result in millions of dollars being spent on process and technology modifications to support it
[12]
"he bill, co4sponsored by 2eps (hris (annon of 0tah and Anna -shoo of (alifornia, was introduced in January =>>>
Add to all this the fact that consumers are feeling harassed !t,s no longer enough to post a privacy policy on a Web site6companies are being blacklisted by consumer organi'ations for contacting customers more freAuently than they,re comfortable with, or for simply sending duplicate mailings "he point to incorporating privacy measures into a (2* initiative is this: the customer doesn,t care about your company,s intent6it,s the !ehavior that counts A duplicate mailing or unwanted phone call can be the innocent result of an incomplete customer profile or simple data error $ut from the customer,s perspective, it could mean all your outbound communications are suspect 0sing a customer,s data without her permission need occur only once before that customer is lost (ompanies are increasingly limiting the number of messages they send to their customers in an effort to maintain good will $ut if your company believes every customer is indeed an individual, contact preferences will be different too 0nderstanding the type of marketing message and its optimal channel aren,t enough6 you must understand each customer,s preferred interaction fre6uenc& as well 3imply put: -ach customer,s privacy preference should be solicited and incorporated into his customer profile and should then be uneAuivocally honored As we mentioned earlier in this chapter, customers have unspoken needs, some of which you,ll have to infer from their past interactions with you Aside from the numerous marketing possibilities offered by (2* and its accompanying customer understanding, the ongoing privacy controversy alone makes integrating detailed customer behavior and preference data worth the time and effort
8iven these three factors, selecting a (2* tool to automate campaign management is easier said than done A good campaign management tool is fle5ible enough to support e5isting marketing processes and not enforce its own !t should also be able to sustain current campaigns, yet support the evolution of the business and its customer knowledge while at the same time leveraging other campaign details such as the campaign,s chosen media, its manager, or its seasonality 3ome of the more advanced products even include so4called decisioning capabilities that support the translation of customer profitability, value, or future behavior scores into automated actions <for e5ample, suggesting the waiver of a service fee for a particular customer? (ompanies in the midst of planning or using (2* technologies to automate marketing are nevertheless still struggling with basic Auestions about how to optimi'e their marketing e5penditures -ven those in the heat of implementing (2* are still asking themselves these Auestions: +ow do we focus our marketing campaigns on customers with whom we,d like repeat businessM +ow do we migrate customers to lower4cost channelsM +ow do other organi'ations in the company see customers differently than we do and how does that influence our campaign messagesM +ow can we anticipate which products and services a customer might wantM What is the best means of communicating with customers on an ongoing basisM What tactics do we use to entice prospects to become customersM +ow do we tie what we,ve learned about customers to improving overall customer satisfactionM What keeps our most loyal customers coming backM
(2* technologies alone can,t help companies answer these Auestions As we,ll see in later chapters, companies implementing (2* must make staffing and technology decisions that transcend (2*
The Challenges&
Despite being ahead of the curve, -ddie $auer is a testament to (2*,s comple5ity With its lofty multichannel and one4to4one goals6goals, the company would point out, that are being achieved6-gler admits they might have been a little too ambitious BWe wanted to get to *ecca right away,C he says +e cites three different (2* tactics: evolutionary, where each step is logical) revolutionary, which involves significant business change) and interdisciplinary, leveraging (2* to a diverse set of decision4makers B-veryone wants to get to ,interdisciplinary, immediately,C -gler e5plains, Bbut there are incremental steps across (2* development We probably tried biting off more than we could chew early on C -ddie $auer has also had its share of business process issues as a result of its (2* strategy, but the resulting changes ultimately contributed to a greater degree of customer4centricity !f he had it to do over again, -gler says he would have made sure organi'ational and business processes were considered earlier in the (2* lifecycle
)ood Ad9ice&
-ddie $auer is intimately aware that the rules of retailing are changing !n these days of hypercompetition, market fragmentation, and !nternet speed, +arry -gler considers a learning relationship with customers to be a cornerstone of the company,s competitive advantage 3uch a relationship begins with the ob#ective of building overall customer value <see Figure 2-9? 0igure 7-A. The learning relationship .Courtes! o" 4ddie Bauer( Inc./
%ike the (2* program itself, such a learning relationship is a never4ending #ourney: "he fact that -ddie $auer has replaced disparate customer data and outdated legacy systems with over a terabyte of online customer information6from names and addresses to itemi'ed past purchases to channel preferences6 makes that #ourney a lot easier
B%ots of (ompanies are "hinking About (ustomer 2elationship *anagement, $ut Progress (an $e &ery 3low6(2* 0nder 3crutiny,C Information Week1 3eptember F@, =>>>
*oney, for one thing According to the same survey, J= percent of the large companies responding planned on spending DF million or more on their (2* initiatives 2elationship marketing tactics such as customer segmentation and automated campaign management pay off only if a company is ready to use the results to improve customer interactions And if a company isn,t clear about its business processes, (2* can back fire Differentiating customer treatment based on partial customer data can result in the wrong message to a customer, ultimately doing more harm than no interaction at all 3uccessful marketing tactics use the results from customer interactions to improve future interactions, paving the way for high4impact decisions such as these: 3hifting marketing dollars toward campaigns more likely to generate high responses 0nderstanding the characteristics of high4value customers, finding such characteristics in customers who have a high value %otential1 and changing interactions accordingly !mproving the effectiveness of high4cost channels <such as face4to4face sales? to ma5imi'e their revenue streams
"he goal of (2* in marketing is to move (4level customers up to $4level customers and $4level customers up to A4level customers and to motivate A4level customers to stay that way6indeed, to buy more !t is to ensure the optimal type and freAuency of communication, regardless of how BstickyC the Web site, the number of free giveaways, or the cost of the advertising campaign !t is to ensure that the company is the customer,s first choice "o succeed on this distinguished mission, a company,s marketing process must be well defined !t must institutionali'e the practice of customer differentiation !t must act on the information it analy'es *oreover, it must not e5ist in a vacuum, but must support the other business processes that surround it, including the customer support and sales processes described in the ne5t two chapters
+afner, .atie, B!s the (ustomer -ver 2ightM 3ervice,s Decline and 7all,C The 5ew 8ork Time"1 July =>, =>>> $arron, .elly, B+urry 0p and Wait,C (or!e"1 /ctober FK, =>>>, p F@E
[2]
"wenty years ago, most consumers complained by writing letters, mailing them, and awaiting a response whose arrival was hit4or4miss "he E>> number sparked a revolution in customer service, offering consumers real4time dispute resolution and the accountability of a company representative 1ow e4business has upped the ante "o its dismay, Dell (omputer recently uncovered the fact that customers who ordered their P(s from the company,s Web site made an average of two and a half calls to the company,s support center [3] <3o much for the Web,s efficiency ? And scathing news reports began citing shoddy customer service as ground 'ero for why online retailers had lost a combined DK F billion in FGGG -4tailers were accused of intentionally omitting contact phone numbers from their Web sites and print ads to avoid costly telephone calls, in effect, forcing customers to interact with a company via its lower4cost Web channel "he fact is that with the !nternet, the call center often represents the customer,s only opportunity to interact with a real human being
[3]
(all centers, increasingly known as contact centers, customer interaction centers, or6somewhat optimistically6customer care centers or even knowledge centers, were around long before customer databases, and friendly service was the hallmark of such companies as the Walt Disney (ompany and 3tate 7arm !nsurance even before the Web made its mark (ompanies have long known that customer support is critical to both keeping e5isting customers and acAuiring new ones What they haven,t always known is how to perform it effectively 0ntil recently, e5ecutives considered their companies, contact centers a necessary cost of doing business and treated them as if they served the sole purpose of weighing down corporate profitability (all center hiring involved merely testing how fast a #ob candidate could type (all center operators had been responding to each call as if it were uniAue, in effect, proving their e5ecutives partially right "he reality is that up to @> percent of call center contacts are duplicates (all center effectiveness was measured by how Auickly the operator could get the customer off the phone When the customer,s call reAuired research, tracking down a product specialist who could work on the problem could take days or even weeks $ecause calls were rarely monitored, some operators took liberties in inventing their own ersat' solutions to turn over as many calls as possible (all center technologies entered the marketplace to effectively alleviate some of the repeat work and increase efficiencies, allowing companies to handle escalating call volumes (all center products offered the creation of trouble tickets for customer complaints and provided tracking of trouble tickets from the seminal call through its resolution (32s <customer service representatives? could look up similar calls and resolutions while a customer was on the phone without having to repeat research Products such as 2emedy and (larify enable companies to run statistics on their calls, categori'ing them by call type, time4to4 resolution, escalation percentages, and average call duration 3uch products also provide forecasting of call volumes to ensure adeAuate call center staffing 3uch vendors legitimately aid their clients in increasing agent productivity and turnover rates, e5pediting training, gauging the effectiveness of solutions offered, and6ideally6shortening problem4resolution times and reducing costs </ne popular ma5im advises that any customer interested in bankrupting a company need merely phone the call center on a regular basis ? *ost companies use statistics from their call center products to track (32 performance, compensating them according to the volume of calls they can support 1otwithstanding consistent improvements in call center software, business processes have still played a larger role than customers do !n Figure 3-1, the customer is not happy "he airline had promised to mail him a ticket, which never came "he (32 is authori'ed to reissue the ticket and log the customer,s complaint %ucky for the customer, the fare hasn,t changed And lucky for the airline, the customer accepts his new ticket +owever, the focus here is not on the customer,s happiness but on the process of ensuring that the complaint is recorded, with the goal of post facto analysis and comparison 3oothing the customer is up to the call center agent6standard operating procedure in most of today,s call centers !ndeed, the customer hasn,t figured into the call center infrastructure at all until very recently
Call Routing
A company,s customer support Bhelp deskC might actually be made up of several6or many6geographically dispersed contact centers "hese contact centers might be segregated based on "he location of the company,s regional offices 3taff e5pertise or product speciali'ation Pro5imity to key research and development <2LD? activities "he location of the firm outsourcing the company,s contact center &arious escalation levels for trouble tickets
$ecause performance remains the central metric of contact center success, minimi'ing the time a customer waits on hold for a (32 stays a top priority "he concept of Bfollow the sunC customer care emerged, meaning call centers were either geographically placed or open around the clock to accommodate customers in different time 'ones (all centers have become like all4night convenience stores: they offer a range of services and are open =H hours "o further minimi'e customer wait times, Bload balancingC <the ability of the network to automatically route a customer,s call to the first available operator? is a critical operational ob#ective 1etwork routers monitor operator availability while keeping track of the incoming calls in the Aueue, using switch technology to allocate calls appropriately "his automatic call distribution not only reduces the time a customer stays on hold by efficiently routing calls to available agents, but it can also apply intelligence about the customer to its decision about where to route a call 7or instance, by using caller identification, a call center system can route a call from a multicultural city such as %os Angeles to a bilingual (32 /f course, companies were routing incoming calls long before ("! technologies automated the process +owever, in these cases, it was freAuently the customer who had to navigate the often comple5 alleyways of a company,s organi'ation *any companies6most notably the airlines6still routinely provide their best customers with a separate E>> number that guarantees they,ll reach specially trained service agents
5eb-based +el"-+er9ice
(ustomers are both refreshed and annoyed by company Web sites re4intermediating the support process After all, the shortest distance between two points is a straight line6shouldn,t you be able to talk to a human being when you wantM "he previously mentioned redundancy of calls to customer support centers has initiated the automation of customer support processes, not only increasing support efficiencies but also allowing customers a greater degree of access to important information via the Web, any time of day, for a variety of Auestions 7or instance, sites such as 7ed-5 com allow customers to track recently called the 7ed-5 support center to track a package "he seconds $y comparison, the 7ed-5 Web site took FJ seconds improvement by the number of people every day who want to know to see why 7ed-5 automated their package tracking their shipments "esting this theory, ! time from initial call to answer was GI from the time of access *ultiply the where their packages are, and it,s easy
$ut customer self4service has as much to do with obtaining general information as it does with tracking specific orders *ost Web sites have made 7AOs67reAuently Asked Ouestions6a site staple 7AOs answer Auestions such as these: Where is the company headAuarteredM +ow do ! return an itemM !,ve moved6how do ! update my profileM +ow do ! review my accountM +ow do ! change my passwordM ! have a Auestion about my bill) what do ! doM
"he more choices customers have in how they can deal with the company, the more likely they will be satisfied with that company,s service "his can mean something as simple as a company e4mail address provided on the Web site so customers can provide the company with details on a Auestion or problem !t can also mean the inclusion of a B(all meC button that allows the customer to reAuest an in4person conversation with a company representative, on the company,s nickel !ndeed, many companies have taken person4to4person electronic support[4] one better by offering customers the ability to chat online with a customer care representative %ands, -nd is famous for this !ts %ands, -nd %iveU service offers customers the capability to ask Auestions of a company rep either by telephone6a %ands, -nd agent will call the customer after she enters her phone number6or by live te5t chat6the customer can engage in an e4mail dialog with a representative <%ands, -nd reps have even been known to ask whether your child has grown a si'e since you ordered those overalls last *arch ?
[4]
!n Chapter 1 we introduced the term Be(2*,C usually meaning the capability to contact a customer service rep through the Web6by e4mail, live chat, or high4speed Web telephone line
B%ive person chatC features bring customer support to real time) customers with a single phone line need not log off the Web site to call an E>> number %ikewise, (32s staffing live chats can service more than one customer simultaneously %ive chat helps ensure that the customer stays on the site and takes his online shopping cart through checkout rather than abandoning it "his increases the likelihood not only of sales, and thus revenues, but of customer satisfaction as well /f course, breakthrough technological capabilities alone won,t ensure stellar customer service 3tudies show many companies still can,t handle the volumes of phone calls they receive A 3outhwest Airlines spokesperson told The Dalla" Morning 5ew" the company would not move toward e4mail support until it could guarantee the same level of service it could provide to its telephone customers <(ustomers calling the airline,s (ustomer 3ervice Department rarely wait more than K> seconds before speaking to a human being, and the company,s Web site is e5emplary for its ease of use ? /ther companies, such as $altimore,s All 7irst $ank, only allow customers in top segments to use live chat features on their Web sites !ndeed, with limited human resources a fact of life for most call centers, companies are constantly evaluating new ways to mitigate customers, need for human contact A recent survey showed that @K percent of all financial services companies either failed to respond to customer e4mails in a timely fashion or never responded at all [5] <"he moral: Don,t introduce new products and features until you,re serious about supporting them ? %ike their telephone counterparts, online customer service reps must have e5cellent communication skills and must be good at multitasking to do the #ob well
[5]
3tudy done by (elent (ommunications, reported by Bri)ge 5ew", August =H, =>>>
3uch electronic surveys are becoming increasingly sophisticated, serving as the technical eAuivalent of the focus group, albeit with more science and sans the overhead (ompanies can randomi'e their surveys and achieve detailed response reports *any companies who might have developed their own Web sites and the accompanying infrastructures have nevertheless chosen to outsource their customer surveys to firms that speciali'e in designing custom Auestionnaires, tallying responses, and mapping survey results to the client company,s overall strategic goals
Call-+cripting
As customer contact centers become more automated, the infrastructure improves (ustomer databases become richer, and customer behavior and preferences can actually be predicted by comparing them to the behaviors of similar customers over time $ecause of this increased customer intelligence, the capability to provide (32s with situational scripts is emerging as a must4have for many contact centers 3uch scripts eliminate agent guesswork by providing the (32 with a logical series of talking points and guiding her through a dialog with the customer based on such factors as "he reason for the contact "he customer,s value (ross4selling opportunities and propensity4to4buy data (urrent product promotions or discounts Past4due bills or accounts payable issues
3ome products actually provide natural language support, meaning the agent can enter a sentence or phrase <B(ustomer needs laundry instructions for flannel sheetsC? and be supplied with the appropriate response <B(old water, tumble dry on lowC? /thers feature logging of transcripts, allowing a company to
!n this e5ample, the angry customer,s profile is maintained on a (2* server or, more likely, on an enterprise data warehouse that contains a superset of information about that customer, not simply his calls to the contact center 3uch a database might include information about the marketing promotions the customer has received, his travel history, how he books his tickets, his seating and meal preferences, and his use of partner products and services !n this case, the customer,s profile indicates most of his ticket purchases are B9C fares, he has flown over K> segments in the past seven months, and he belongs in the top I percent of the airline,s freAuent fliers "his information in turn dictates a script that guides the (32 in offering the customer a free upgrade along with his overdue tickets "he customer is more likely to remain loyal and consider the airline,s error an anomaly rather than a reason to churn "he ability to differentiate treatment of different customers based on their value is not only the +oly 8rail for contact centers) it is the cru5 of (2*
C!beragents
3ome (2* vendors are taking the concept of self4service one better by introducing cyberagents, lifelike BrepresentativesC normally depicted on a company,s Web site as a real person6(omp03A,s BJillC being one of the archetypes (yberagents attempt to pull together the best of both personali'ation and advanced technology /n the one hand, the cyberagent is given a personality6complete with voice and facial e5pressions6often
5ork"orce #anagement
3taffing the contact center has traditionally been a matter of determining who can work when "he contact center with three people dedicated to the "op F>> customers, si5 more covering the ne5t tier of @>>, and so on, is bound to fall short of customers, e5pectations, not to mention employees, !t,s great that the best customers are getting better service, but is it good enoughM Workforce management tools speciali'e in staff planning and optimi'ation, and several products are specific to contact center staffing /ptimi'ing staff around high call volumes, different communication channels, and customer types is the surest means of guaranteeing that the right customers receive the appropriate levels of support 3uch products can combine operational contact center tasks, such as call routing, with planning functions including Ability to forecast contact volumes to predict busy periods 2ecommendations for the optimal number of (32s for certain peak periods
Performance tracking by customer value, customer satisfaction, priority level, or other metrics -mployee scheduling based on skills, tenure, or preferred work hours 8lobal monitoring of multiple contact centers, with the ability to combine findings into single reports for staff or performance fine4tuning
With staffing issues claiming up to J> percent of a contact center,s operating budget, the ability to track performance against customer satisfaction levels and fine4tune staffing choices accordingly is the goal of
+cenario
!t,s 7riday morning and you,re at home waiting for the D3% rep to install a third line, so you,re teleconferencing into your biweekly staff meeting -ach of the eleven team members gives a brief status about his or her department Just as you,re about to report on profitability for OH, your second phone line rings "hinking it,s the D3% rep6finally you can nail down a time:6you beg forgiveness, put your waiting teammates on hold, press line =, and hear a friendly voice B+i there::: !t,s (olleen:C says the voice (olleen is warm and enthusiastic, and it takes you a minute to reali'e you don,t know a (olleen $y this time, (olleen is gushing about the fantastic prices on gemstone #ewelry that will be featured this weekend on home shopping television 9ou hang up and return to your meeting BWho was thatMC an indiscreet staff member asks B(olleen,C you say in spite of yourself, unknowingly raising more Auestions than you,ve answered When the conference call wraps up, you call the home shopping channel and reAuest to have your phone number removed from their automatic telemarketing list "he operator is pleasant and helpful and promises he,ll remove your name from the mailing list +e reAuests your e4mail address, and you decline 9ou,re cleaning up your life, getting rid of unsolicited marketing messages from companies you don,t care about 9our call to the home shopping company eliminated one more unwanted phone call 9ou hang up, feeling you,ve accomplished something A few hours later, the D3% rep calls +e,ll be showing up atP "he second line rings 9ou interrupt the D3% rep and take the call !t,s (olleen again 3he,s perky as ever, this time flogging an upcoming kitchen appliance show, and you now hate her guts !t,s no surprise that, in spite of revolutionary advances in call center software functionality, established industry best practices have nevertheless failed to promote the infrastructure necessary to support evolving customer needs !ndeed, according to one recent survey, not even technology companies are immune: *any dot4coms also actively avoid the considerable investment necessary to build an adeAuate customer service center to answer phones and e4mails from disgruntled customers [6]
[6]
/lsen, 3tefanie, B(ustomer Disservice: -4tailers Dodge (alls to (ut (osts,C (1-"1ews com, 3eptember FH, =>>> <www !"et"ew# !$%?
3o what can your company do to ensure your call centers are able to respond to customers in the optimal way, recogni'ing who they are and providing a consistent level of serviceM 7ollowing is a checklist for customer support success: hoo"e &our technologie" carefull&* 0nderstand your business needs before you invest in an e5pensive and comple5 call center system and all its e5tensions 7or instance, if your company,s Web site isn,t tied to inventory yet, it makes no sense to offer customers Web4based self4service and risk unmet e5pectations %ikewise, if your company,s most valuable customers land in voice4 mail hell at every contact attempt, something,s wrong +ave the call center evolve with the business, which might mean a building4block approach to implementing your call center infrastructure Just be sure your call center technology is a step or two ahead of your business, evolution so there are no awkward surprises 'rovi)e /R" with ever&thing the& nee) to know a!out cu"tomer"* Although this metric implies the integrated customer data from a variety of contact points <see Chapter 6 for more on this?, it also mandates that a company be aware of what various call centers might reAuire so they can provide the best service 7or instance, if past4due customers are routed to a collections operator, that
/taff the contact center at a%%ro%riate level"* Web shoppers reAuesting live customer support routinely receive messages that say something like this: B"here are currently no agents available to assist you (lick on the Auestion mark below to send us an e4mail C !f your company is going through the e5pense of providing multichannel customer support, it should be willing to optimi'e support resources for individual channels 0se e5isting call center software to monitor and track call volumes according to time of day and channel and then plan staff accordingly 7or more comple5 environments, find a workforce management tool that can help you take the guesswork out of staffing your call centers /hare ke& learning"* After all, customers don,t call the marketing department when they need help 3haring contact center activity reports with other areas of the company can aid the company not only in better understanding its customer base, but in improving its products as well 7orward4 thinking companies with centrali'ed customer data actually enter important facts learned from customer interactions into those customers, profiles, ren dering the information available to the company at large When the right processes are in place, redundant trouble tickets and freAuently logged complaints can make their way into 2LD, resulting in product improvements and differentiation6Bclosing the loopC on high4profile problems Im%rove /R com%en"ation* After all, how much can someone earning a low hourly wage realistically care about an unhappy customer, particularly if she,s #ust been trained to cross4sell products and survey customers in addition to solving their problemsM (ustomer service is stressful, rarely rewarding, and e5tremely visible $alance (32 compensation with realistic business
The Challenges&
"he lu5ury of being able to build a company,s support infrastructure from scratch is also its difficulty "rue, Juniper has no neighborhood branches with lines of impatient customers spilling out the door 1or is there an easy way for Juniper to provide its customers with an intimate human e5perience "here are no unwieldy legacy systems to grapple with, but that means little established data to direct the best decisions BWe were always tempted to revert to our past e5perience,C Wallace reflects, Band sometimes we had to guess $ut our products are different, and so are our customers C All things being eAual, Wallace estimates Juniper is between F= and FE months ahead of the industry curve simply because she didn,t have to undo e5isting business processes to establish effective ones 3he plans on tackling the organi'ational issues ne5t: BWe,re moving toward organi'ing more around the customer so that we can continue to evolve our customer focus C
)ood Ad9ice&
Wallace also credits her success to the tight relationship between customer support and marketing "he classically tense relationship often sees service reps accusing marketing of being unrealistic about its promotions, and marketing accusing customer service of being unresponsive and foiling its campaigns 1ot so at Juniper, where product managers in marketing can regularly monitor how their campaigns affect inbound customer contacts, and 2*s can openly provide feedback to marketing about campaign improvements "he two organi'ations meet regularly and maintain a weekly performance scorecard "he organi'ational missions of both Juniper,s (ustomer 3upport department and its *arketing department are centered on the customer e5perience As Juniper understands more about its customers, behaviors and preferences, 2*s can influence them 7or instance, if a customer purchases a Juniper credit card through a direct4mail promotion, an 2* would highlight the benefits of !nternet banking in his welcome call to that customer, potentially steering her toward a more profitable channel and evolving her toward online bill payment !ndeed, when Juniper communicates with a customer on his palmtop, reminding him that his telephone bill is due and offering to pay it for him, the line between marketing and service becomes refreshingly blurred
*oreover, the scenario assumes that headAuarters possesses the necessary information and can communicate it Auickly6many didn,t and still can,t "he promise of sales force automation was in putting account information directly in the hands of field sales staff, making them responsible for it, and ultimately rendering them <and the rest of the company? more productive 37A likewise helped managers in both the front and back offices who, lacking timely sales forecasts, struggled to keep up with demand /rgani'ations needed a way to be sure that up4to4date sales data, including sales forecasts, could be communicated across the organi'ation from the manufacturing floor through to company e5ecutives *anagement considered the potential for sales force automation to increase productivity and provide consistent information as nothing short of a competitive weapon 3alespeople took a little more time to warm up to the idea "o be fair, some are still warming up 3ure, they could now keep track of product inventories, customers, names and addresses, org charts, and marketing
Toda!'s +0A
37A product vendors set out to automate the answers to some basic sales rep Auestions that, when answered, could help them be more productive 3ales force automation products run the gamut from rudimentary calendar support and scheduling to real4time alerts about customer events !rrespective of the vendor or product, most sales force automation tools claiming the (2* moniker fall into one or more of the following categories
Contact #anagement
(ontact management is the subset of sales force automation that deals with organi'ing and managing data across and within a company,s client and prospect organi'ations (ontact management software can contain various modules for maintaining local client databases, displaying updated organi'ation charts, and allowing salespeople to maintain notes on specific clients or prospects *any also allow users to Auery remote databases for supplementary information or to synchroni'e laptop4local data with a corporate customer database With contact management tools, salespeople can answer the following types of Auestions: Who is our contact individual for (lient NM Who is the contact for each departmentM
*uch of the (2* software available today that falls under the contact4management rubric interfaces with *icrosoft /utlook or similar products "his linkage allows sales reps to supplement their phone lists and calendars with fulfillment features to track customer mailings or automated workflow programs that can assign and route appointments As part of a larger (2* product suite, many contact management tools enable salespeople to communicate their schedules to the organi'ation at large "he real value of contact management (2* is in its capability to track not only where customers are but also who they are in terms of their influence and decision4making clout *any contact management tools can interface with sales management functions to help optimi'e the number and skills of team members involved in high4profile sales efforts (2* suite products can then leverage these features to map sales activities back to resulting revenues
,ead #anagement
(2* vendors claim that the capabilities described in the previous section push sales prospects through the pipeline more Auickly $ut who,s in the pipeline, anyway, and how long will they be thereM %ead management <also known as Bopportunity managementC and Bpipeline managementC? aims to provide foolproof sales strategies so no sales task, document, or communication falls through the cracks "hus salespeople can follow a defined approach to turning opportunities into deals *any lead management products not only track customer account history but also monitor leads, generate ne5t steps, and refine selling efforts online 3ome sales force automation products enable a company,s marketing or sales management organi'ation to automatically distribute client leads to a field or telemarketing rep based on the rep,s product knowledge or territory $ecause the organi'ation has presumably already Aualified these leads through marketing campaigns or client referrals or via an internal customer support rep, this type of lead distribution can result in faster deals and higher close rates !n addition, 37A tools can track other prospect attributes such as known product interests, discretionary budget amounts, and likely competitors, providing a real4world view of each lead and its likelihood of becoming a full4fledged sale "his information can feed sales forecasting software and result in more accurate predictions than relying on traditional spreadsheets of past performance /ther 37A products bring the sales process full circle, tracking leads against orders to provide a view of close rates and salesperson productivity "his information can then be reported back to the marketing manager who originally designated the lead "his enables marketing to compare a campaign,s result to actual sales of that product and refine future campaigns over time [1]
[1]
"he ability to record campaign results and leverage them for future campaigns is one e5ample of Bclosed loopC campaign management
3uch capabilities can result in answers to Auestions that previously depended on guesswork or analysis of incomplete data, for e5ample: At which step in the sales cycle do we lose most of our prospectsM +ow many sales appointments did Adrianna have with "ech(oM +ow many were cancelledM What percentage of leads resulted in sales in the northwest region last monthM +ow long is the average sales cycle overallM +ow does that compare to the sales cycle for the health care industryM +ow did order amounts for the 3an 7rancisco branch office compare to those in 3eattle last yearM
/f course, lead management analysis is only as good as the data that feeds it, and different products provide different levels of sophistication Advanced lead management tools even calculate the probability of
Con"iguration +upport
$ecause many of these tools allow a salesperson to input client and prospect information into an easy4to4 use tool, (2* products have evolved to leverage this information by providing product4specific configuration support to companies who must BbuildC products for their customers 3uch companies6computer technology vendors, appliance manufacturers, and telephone companies among them6no longer undergo the painstaking process of factoring in comple5 customer attributes and reAuirements to build a solution from scratch (2* vendors providing this type of functionality often use the graphical sales process like the one pictured in Figure 4-2 as a way to consider each step of the cycle After the BorderC stage has been reached, the tool can calculate a product configuration and price Auote automatically When this is finished, a configuration tool can provide forms that facilitate electronic communication of the information to other areas of the company 7or instance, an account manager can generate a configuration and price, e4mail his district manager the price Auote for approval, and check inventory for stock, all while "itting with the cu"tomer* 3iebel (onfigurator from (2* heavyweight 3iebel 3ystems also supports headAuarters marketing staff and product managers who might create new product bundles for the field !nstead of field salespeople combining products and selling them as ersat' units, marketing staff can use the (onfigurator tool to build uniAue product packages, brand them, and distribute them to field sales staff with a few mouse clicks After the product has been configured and a Auote calculated, the information can automatically populate a standard contract resident on the salesperson,s laptop 3uch automation dispenses with paper4based contract generation and validation steps so fraught with error they can bog down or even kill an order
Fnowledge #anagement
As field salespeople know all too well, a lot more than customer data is necessary for selling, and the more information available, the better /rgani'ations have a plethora of information an account rep can use sometime during the sales lifecycle Accessible internal documents can provide the sales force with the information it needs to understand a variety of components in the sales lifecycle 3uch information might include (orporate policy handbooks 3ales presentation slides (ompany phone list Proposal templates (ontract boilerplate -5pense report forms 2egulatory standards and recent compliance reports +istorical sales and revenue reports Partner and supplier profiles "ranscripts of sales meetings and e5ecutive briefings Digiti'ed video of sales presentations or e5ecutive briefings !ndustry and competitor data 1ews articles and press releases "rade show and promotional event schedules "hank you notes and other client correspondence
+ow effectively salespeople use this disparate information depends on how easily they can access it *ore and more companies are developing corporate intranets for headAuarters and field staff alike, facilitating the dissemination of critical corporate knowledge
Although not e5clusive to sales organi'ations6knowledge management systems are increasingly popular with large companies such as consulting firms that have geographically dispersed employees6the e5istence of a consolidated view through which the field force can access a variety of information delivers increased efficiencies [2] 1o more mailing binders to the entire 1orth American sales team: %et,s give them access through the .* system: .nowledge management can mean the difference between losing and retaining key sales staff who are hungry for information but strapped for time
[2]
"he documentation and information resident in knowledge management systems are usually displayed via an enterprise portal, a type of BwindowC into the system that makes the information appear centrali'ed
+andheld device technology is evolving at a fast clip According to the Aberdeen 8roup, JH million people will have access to the Web via wireless technology by the year =>>H Personal digital assistants <PDAs?, cell phones and Web phones, two4way pagers, tablet P(s, and any other device with a display screen and the appropriate communications features can support anytimeTanywhere access via wireless networking "he benefits are boundless Deals can be won or lost based on who can respond the fastest With wireless technology, configuring a product, pricing it, and creating the contract can all be done remotely and in much less time than by traditional fa5 and e4mail based processes 1ot only can field staff access important information, they can also enter updates to that information that are relayed back to a corporate server accessible by other organi'ations (ompanies get smarter about their customers as they witness the creeping customer intelligence that accompanies such a closed4loop information e5change As in 3cenario = at the beginning of this chapter, an account rep can be alerted about in4progress customer events in real time 3he can then send a message via a two4way pager or other device to a field service rep for an e5planation, to her boss for advice on how to handle a touchy client situation, or to her marketing contact for up4sell suggestions Wireless messaging can be invaluable when it comes to #ust4in4time personali'ed messages "hese BpushedC messages6meaning they are sent unbidden but nevertheless geared to the user,s stated interests, subscriptions to news services and other information providers, and even geographic location6can include broadcasting traffic bulletins en route to a client, automatic calendar updates, and daily company press releases Personali'ed messaging also mitigates the BnoiseC accompanying the ascent of handheld communications by limiting information to only those messages the user has reAuested %ast but not least, wireless !nternet access provides field staff with access to vast amounts of information, restricted only by the capacity of the device to display the data 1ot only can a sales rep view e5ternal Web pages on his PDA to glean data about clients, stock prices, and business news, he can also log on to his company,s intranet, thus accessing company4specific information in real time from wherever he is An account rep on the way to a sales meeting can check inventory for a product at an airport !nternet kiosk before boarding his flight -merging technologies such as the so4 called voice browsers even allow users to navigate the Web via vocal commands over the telephone "hus a sales rep can make a phone call during the seventh inning stretch to verify a client address, check on the status of his customers, orders, read his e4mail, and order flowers for 3ecretary,s Day
&arious technology developments are responsible for bringing information literally into the hands of disparate business organi'ations and communities that nevertheless must e5change information on the go +aving emerged to great fanfare in the early FGG>s, the Java programming language meant applications could run on different platforms N*% <e5tensible markup language? simplifies intersystem communications by encoding descriptive information about that data so it can be gathered and shared across applications and companies 7or e5ample, N*% can define information about a purchase order so when a salesperson transmits it, the company,s accounting system can recogni'e and process it 1ew wireless protocols such as WAP6a set of standards for the transmission of content to different handheld units6allow information to be conveyed to a variety of devices in real time "hus, a company,s lead management system can transmit sales leads to a salesperson,s cell phone for Auick turnaround $luetooth is the new standard for short4range wireless communications <it allows laptops to communicate with tiny printers no larger than a rib eye steak, and it,s the way those handheld computers buy sodas from vending machines? And emerging Web portal software provides your cell phone or PDA with a Windows4 like interface Palm recently launched its *yPalm portal, allowing users to wirelessly synchroni'e information on their Palm devices with information on their desktop P(s 7orgot what time the $ig *eeting starts and can,t get to your office to look at your online calendarM Access it right from your Palm &endors are providing a variety of choices for real4time access to their centrali'ed (2* servers with products designed for specific access devices *any (2* vendors offer complementary products for the Web, handheld devices, and thin clients alike !ndividual salespeople can select their preferred way of
"he software infrastructure reAuired to support mobile applications is #ust starting to become available 7or instance, many 37A tools don,t provide their own document management systems, so things like making contracts and other documents available to wireless clients within 37A systems can still be impractical "hankfully, there are tools6like 3un /1Web"op6that allow people to maintain various files and access them from any device !ndeed, mobile computing means new users, new devices, and new ways of doing business 0ntil companies get their arms around the mobile business model from both infrastructure and cultural perspectives, the laptop will live on
1ote that each communication involved in Figure 4-3 can be performed using wireless technology, from the customer reporting the problem via his cell phone to the field service rep closing out the trouble ticket via wireless laptop (onsidering that only five years ago most of these interactions took place on paper, these processes have become <and continue to become? e5ponentially more efficient, generating not only faster turnaround times but also higher customer satisfaction rates "he (32 can actually enlist the support of product specialists or field service engineers by sending a message to their wireless devices With the right analysis capabilities, the (32 can rate the severity of a customer problem and then determine whether a field service rep should be dispatched, based on either her particular skill set or her current distance from the customer,s site !n this way, the (32 can dispatch the technician most likely to fi5 the problem the fastest 7ield technicians not only receive dispatch orders via their PDAs, pagers, and cell phones) they can even make use of these devices during the actual repair "he ability to pop a (D into a (D42/* or insert a memory card containing installation instructions, repair guidelines, or eAuipment specifications provides the technician with the information he needs while at the customer site and frees him from lugging around bulky how4to manuals "he technician can use the device to order additional parts and supplies, as well (ommunication software can then alert the (32 handling the trouble ticket that additional eAuipment is needed Figure 4-4 illustrates some of the features inherent in the newest wireless devices, some of which look and work more like micro4laptops than like the pagers of old 0igure ;-;. A "ield ser9ice wireless terminal
"he advanced features they provide sure beat the alternative: the technician trying to find a computer someplace so he can dial a corporate server or access the Web to order the part and enter the updated repair status6leaving the company and the customer at the mercy of a single field rep,s busy schedule !n real life, field engineers for my local telephone company send completion status of their fiber4optic installations to the company,s large customer database via *otorola =4way pagers, ensuring near real4time repair updates 1o more calling the home office with their whereabouts) no more syncing of databases *oreover, (32s and sales reps can interact with the same data in real time Although current (2* products remain more sales4oriented than service4oriented, customers, field service reAuirements are nevertheless more Auantifiable6and higher4profile6than their sales needs, and much more time4critical Additionally, 77A comprises a set of customer touchpoints that should be recorded as part of the customer,s profile, so when the account rep in 3cenario F prepares his ne5t pitch, he has timely information at his fingertips !t,s no wonder field force automation is the fastest4growing area of (2*
(ommunicate the value proposition to the sales force up front *any6arguably most6first attempts at sales force automation fail because using the tools provides no apparent net benefit (ommunicate the benefits of automating key tasks, defining repeatable processes, integrating disparate data, and knowing customers better as they pertain not only to the sales force6as in, B9ou no longer have to run and mail in sales reports) management can now do it themselves:C6but to the entire company !f possible, Auantify the value of these benefits in terms of potential revenue
"he range of customer touchpoints availed by a robust sales force automation infrastructure can mean the difference between sporadic name4and4address updates and an efficient flow of communication between account teams, their clients, and the company at large
The Challenges&
/verly is frank about his biggest challenge so far, and it,s no surprise: deploying (2* worldwide has its trials 1o matter where they are in the world, sales reps share the goal of being as productive as possible) however, every salesperson has a preferred way of getting the #ob done /verly and his team have had to surmount habits and assumptions that are often not only organi'ationally entrenched but also cultural BWe,re no longer talking about every country,s having a uniAue personal productivity tool,C he says, Bbut about an +P solution C $ecause /racle,s (2* environment is based on portal technology, salespeople can customi'e their interfaces to include client information and contact activities as well as data from e5ternal news sources and the global financial markets "he company is delivering much of the training via the Web to educate staff on how to use the /racle (2* technology and apply it to their specific #ob functions Although most of the training is conducted in -nglish, the company offers locali'ed Web courses in native languages to several countries, including .orea and Japan Another challenge has been rendering ongoing management sponsorship and involvement as painless as possible for +P,s busy e5ecutives BWe reali'ed early on that we,d need to make it easy for e5ecutives to support us,C claims /verly, whose team includes a (2* (hange *anager $esides establishing ongoing (2* performance measures and training, the (hange *anager is chartered with developing and tracking all pro#ect communications, including internal communications across +P, such as guiding e5ecutives in communicating new customer4focused policies
)ood Ad9ice&
7rom the beginning of the (2* initiative, /verly,s team shrewdly steered away from the point4solution approach, preferring instead to e5pend the e5tra time, money, and internal education necessary to deploy enterprise4wide (2* "his meant organi'ing (2* development around so4called Bvertical silosC representing corporate functions such as internal sales, marketing, and customer service "he (2* team members for each vertical silo work with the e5ecutives in related business organi'ation to design the future4 state business processes !n addition, a hori'ontal design group is responsible for the cross4functional process design
3ays /verly, B-very day there,s a new challenge6you need someone who has a good understanding of the problem and the authority to fi5 it C
As reported in B+ow .o'mo is 8etting .illed by its (ustomers,C The 5ew 8orker1 3eptem ber H, =>>>
eCR# 49ol9ing
What about the loyal .o'mo customer who spent an average of DF>> per week and wanted that Dove bar without the surchargeM With the right (2* functionality, a delivery service would know when to waive the surcharge to keep the customer !n many such cases, F> percent of a company,s customers account for G> percent of the profit 0nfortunately, dot4com companies offering commodity4based products have been investing more in their basic infrastructures6warehousing and distribution, recruiting costs, and such6and less in analysis capabilities that would tell them who their good customers are 3ome would argue they have no choice /thers would maintain that without a clear view of the customers they want to attract and those they want to keep, the best logistics and the best procurement in the world won,t matter Pure4play dot4com retailers are getting a run for their <dwindling? money these days from traditional retailers who have parlayed their e5isting infrastructures into a competitive edge After all, brick4and4mortar businesses have mastered economies of scale, having learned the lesson early that a wider variety of well4 priced products will draw more customers, and they are e5pert at getting rid of nonproductive inventory !n leveraging the !nternet, these brick4and4 mortar firms find themselves firmly planted in the world of e4 commerce, with the systems and processes in place to support them
#ultichannel CR#
3ome predict that the brick4and4mortar companies will have the last laugh After all, for every futurist proclaiming the Web will revolutioni'e the world, there,s a pragmatist who pronounces it B#ust another channel C Although both might be right, it,s where the !nternet provides the most value that it incites the greatest change (ompanies such as high4end retailer "iffany <www tiffany com? simply use their Web sites as an additional means of luring people into their stores by providing add4on customer service and product4related information on, for instance, choosing a diamond 3imilarly, %evi 3trauss L (o includes a retail store locator on its %evi,s <www levi com? and Dockers <www dockers com? Web sites so customers know where to go shopping $loomingdale,s <www bloomingdales com?, with limited physical stores across the 0nited 3tates, has dramatically e5tended both its presence and its brand through its Web site, allowing its Web shoppers to purchase their items by referencing the product,s catalog item number "he Web has become a cost4 effective way of increasing the corporate footprint !ndeed, many acknowledged e4commerce best practices are retailers6like -ddie $auer, profiled in Chapter 26who combine powerful online features, including customer self4service such as order and shipment tracking, with an established brick4and4mortar presence, giving customers the e5perience they still want "he award4winning J (rew Web site <#crew com? offers more than #ust apparel: speciali'ed services uniAuely available in cyberspace J (rew Web shoppers often see promotions before they make it into the
Petco <www petco com? offers pet lovers a community center where they can chat with animal behavior e5perts and access a range of articles "he site also features an online Bpet yellow pages,C where visitors can find a range of pet4related services available near their homes, and even provides a way to donate money to the A3P(A online And drugstore com <www drugstore com? features the ability to analy'e the interactions of certain drugs, and the company,s e*edAlert program provides daily information about drug product warnings and recalls $ut such value4added features motivating customers to return to the Web site, catalog, or store don,t necessarily guarantee customers will buy more of a company,s products -ffective marketing means more than #ust a slick Web site and automated campaigns And the more companies know about how their customers prefer to interact with them, the better service they can provide
CR# in B7B
(ompanies that are actually communicating with online markets have flung the doors wide open "hey,re constantly searching for solid information they can share with customers and prospects via Web and 7"P sites, e4mail lists, phone calls, whatever it takes "hey,re not half as concerned with protecting their data as with how much information they can give away "hat,s how they stay in touch, stay competitive, keep market attention from drifting to competitors 3uch companies are creating a new kind of corporate identity, based not on the repetitive advertising needed to create Bbrand awareness,C but on substantive, personali'ed communications [2]
[2]
%evine, 2ick, et al , The luetrain Manife"to$ The En) of Bu"ine"" A" 9"ual Perseus $ooks, FGGG
"here,s no Auestion that the !nternet has streamlined business processes, providing efficiencies unmatched by traditional ways of doing business "he Web makes things happen faster $ut is speed enoughM
*a#or -2P vendors such as People3oft and 3AP not only automated these and other functions) they also linked them for companies who previously had disparate systems that had never been interrelated "hese companies replaced their outdated legacy systems and enabled integrated operations across the enterprise "he products themselves reAuired hefty investments, often into the millions of dollars, and -2P implementation resources usually doubled those budgets "he integration piece alone was a boon to companies With -2P, salespeople could access a single system to check inventory, a purchasing agent could look up a supplier,s pricing history, and a marketing product manager could track defects Despite war stories about underestimated budgets and overestimated consultants, -2P delivered across4the4board efficiencies "his integration had dramatic effects on downstream customer4facing business processes 2avi .alakota and *arcia 2obinson describe (olgate,s -2P success in their book e#Bu"ine""$ Roa)ma% for /ucce""$ $efore 3AP 2TIPdistribution planning and picking used to take up to four days) today it takes FH hours !n total, order4to4delivery time has been cut in half [3]
[3]
.alakota, 2avi and *arcia 2obinson, e#Bu"ine"" 2*<$ Roa)ma% for /ucce"" $oston: Addison4 Wesley, =>>>, FEI
"he supply chain increasingly relies on the communication of customer behavior data <purchase patterns, in particular? to render production scheduling more efficient and intelligent than ever $ut first, a company must culturally believe that sharing information with its suppliers is important !f it is fundamentally reluctant to reveal its sales prices to its suppliers, all the (2* in the world won,t matter After all, there might be sensitive competitive information "he company might be willing to let its supplier know how many product units have sold but nevertheless not want to give the supplier the specific sales price and risk that supplier,s reveal ing this price to the competition 3uch companies must consider that already a wealth of information is moving back and forth between themselves and their suppliers6additional information might be added over time as a means to further reduce costs and enhance productivity +ow do companies know which suppliers to selectM +ow do they Aualify the best choicesM 3upplier relationship management <32*? products help companies analy'e vendors based on whether they are considered strategic suppliers6those with which the company has established a mutually beneficial collaborative strategy6or commodity suppliers, which the company selects based on price (ompanies are leveraging the !nternet more than ever to Aualify, track, and monitor their supplier relationships 32* tools such as the one from 3A3 allow companies to assess the relative cost, value, Auality, reliability, and risk of individual suppliers, thus optimi'ing supplier Aualification and selection %ikewise, suppliers are motivated to improve their relationships with their business customers !n the world of e4commerce, the supplier intent on providing its customers with the eAuipment it needs, when it needs it, is the supplier who gets points for service as well as price "he ability to respond to6indeed, to predict6 product demand means being able to intermingle forecasted orders with actual orders *any companies have posted their current demand forecasts on private portals where suppliers can access them at will 3uppliers, too, are getting smarter $ased on their own abilities to forecast demand for eAuipment, they can participate in the growing number of reverse auctions6in which buyers hold online bidding wars in search of the supplier offering the best deal6with optimal pricing, often in real time *oreover, with newfound capabilities enabling them to analy'e both their own and their customers, data, their bids can be more precise 3uppliers are increasingly digiti'ing their product catalogs and establishing e4commerce infrastructures in line with their customers, improved supply chain management capabilities 3uch collaboration means easier and faster access to products both internally and e5ternally, as well as supplier differentiation in an increasingly commoditi'ed environment *any claim this type of $=$ interaction between suppliers and their commercial customers to be the precursor of the business4to4consumer (2* model that has claimed the lion,s share of the market thus far "he line between $=$ and $=( (2* is becoming increasingly fine As multichannel (2* takes shape, both
3ome P2* products also help companies Aualify and recruit new sales partners <Prior to P2*, Aualifying sales partners involved in4person meetings, philosophical discussions, and tons of paperwork ? Products like these help companies establish desirable partner attributes and allow them to assign partner categories that might limit partners to specific product assignments or sales strategies <see Figure 5-2? 0igure <-7. Channel5a9e's %R# %roduct
P2* tools also provide partner profiles that enable a company not only to understand a business partner,s characteristics6from their headAuarters address to which competitive products the partner might be selling
$e it tighter supply chain integration, optimi'ation of alliance partner ships, or customer self4service, understand (2*,s ob#ectives before deciding on its mechanics <We,ll see a before4and4after business process scenario in Chapter 7 ?
Table >-7. 0our CR# %erspecti9es( 0our 'i""erent 4 ecuti9es E$ecuti%e &'a e( on CRM &P of *arketing BWe desperately need a (2* system We don,t even know how many real customers we have: !,ll pay for the darn tool, but we need it fast !,m thinking of a customer profile dashboard 1ow, when can we get startedMC &P of 3ales B/f course we need it, but we,re not ready yet: !,ve #ust funded an effort to clean up all our contact data After all, what good is customer profiling if the data makes no senseMC
&P of (ustomer B(2*M Why, we,re already doing it: /ur call center system gives our service reps customer 3upport information automatically when the customer calls in We can even see a list of products for that customer: Who says we need anything elseMC &P of -ngineering B3ure, customers are important but it,s taking us K@ days to provision new services !f we don,t fi5 that first, we,re not gonna have any customers left: !f we only knew how to prioriti'e our installationsPC
After all, the ma#ority of (2* products started out as so4called point solutions, designed to solve a specific business function such as sales force automation -ach of these products typically used a local database to store current customer information, hardly a flawed architecture $ut the proliferation of these (2* tools around the enterprise begat assorted and mismatched customer files, all of which were critical to their respective systems, but none of which were linked: 1otice in Figure 6-2 that each database contains key customer information critical to the purpose and functionality of the (2* system "he customer data in each database is different, depending on the reAuirements of the organi'ation using its data $ut as in real life, some organi'ations need the same data about customers) thus the 37A and contact center databases both contain customer contact information as well as payment status, and the 37A and marketing databases each store sales revenue data "hough the names are the same, the contents and format of the data elements themselves are likely to be drastically different across systems 0igure >-7. 2ne Compan!( #ultiple Customer 'atabases
B-5ternal dataC means data coming from outside the company "his can include anything from competitive data from a market research firm to consumer change4of4address data from national data service firms
"he greater the number of sub#ect sources for a data warehouse, the richer the information available to the business Figure 6-3 shows the type of data that typically moves to and from a data warehouse to provide an integrated view of the customer 0igure >-:. Integrated customer data on a data warehouse
/ne of the strengths of a data warehouse is in its ability to store large Auantities of historical data, enabling companies to compare customer behaviors over time 7or instance, by storing customer purchase history, a company can evaluate what might have attracted a customer to making a purchase or gauge whether that
(learly, integrated customer data is mandatory for a company to serve its customers well *otivating the customer to come back6remember our definition of (2* in Chapter 1M6means understanding more than a customer,s name, current address, and income level !t means knowing his preferred products, his consumption rates, values, lifestyle, life stage, and even a superset of his behaviors out"i)e of his relationship with your company And it works the other way: customers respond better with a unified view of the company "he scenario in Figure 6-1 can deliver disastrous conseAuences in the all4too4common instance in which the Web shopper and the customer reAuesting assistance are in fact the same person 2eceiving a cross4sell promotion a day after complaining about a product might irritate a customer on the brink of leaving B! #ust complained about the product and now they,re trying to sell me another onePdidn,t they hear meMC is the likely refrain from the customer who,s received dis#ointed communications from different organi'ations within one firm
"he customer information that can provide answers to these and other business Auestions can drive key decisions about customer treatment, sales techniAues, and upcoming promotion strategies, among others $y integrating operational (2* data with information from around the enterprise, companies can begin performing analytic (2* and, with it, make truly customer4centric business decisions "he practice of data analysis predates even databases and transcends data warehousing and (2* Plus, it,s a varied discipline ranging from standard Aueries to statistical analysis to comple5 predictive modeling *any (2* vendors have incorporated analysis6thus the term Banalytical (2*C6into their products, thereby offering users the ability not only to perform key (2* business processes but also to apply business intelligence to these (2* functions to make them more accurate and more valuable 7or instance, - piphany has incorporated predictive modeling into its toolset to provide lists of customers most likely to respond to a given marketing campaign /ther (2* product vendors have incorporated purchase4pattern recognition into their offerings, enabling marketing and sales staff to compare customers with like behaviors so they can position new products to an optimal audience
"he main difference between operational and analytical (2*, as introduced in Chapter 1, is that analytical (2* is the only means by which a company can maintain a progressive relationship with a customer across that customer,s relationship with the company "his means being able to track a range of customer actions and events over time, using data from operational (2* systems as well as from other enterprise systems, as shown in Figure 6-4 0igure >-;. Anal!tical CR#& The sum o" its parts
Analytical (2*,s IK>4degree view emphasi'es communicating with customers across channels based on the optimi'ation of relevant two4way and personali'ed interactions, be they through a new marketing campaign or from a caller,s complaint history "his offers the ability to transcend more broad4brush customer segmentation and deploy customer communications that are truly one4to4one (ompanies with both operational and analytical (2* capabilities are changing business strategies to 2eward customers with personali'ed discounts and perks for using lower4cost channels Proactively offer products and services that fit a given customer,s needs based on what the customer has already purchased !ncrease purchase rates by dynamically personali'ing content based on the Web visitor,s profile Ad#ust per4customer marketing e5penditures based on lifetime value scores Analy'e combinations of touchpoints across channels to predict a customer,s ne5t likely purchase 2elate high Web traffic to individual visitors and customer segments to better understand Web use and improve Web design "ailor commissions and incentive programs for sales partners based on the value of the customers they bring Prevent a customer from churning by offering incentives based on individual preferences
"he advent of (2* suites has taken up a lot of the slack when it comes to data integration 0ntil recently, companies that didn,t already have established data warehouses were out of luck if they tried to integrate disparate (2* products, software, processes, and data "he #ob was #ust too difficult 3uites6single vendor solutions offering a range of functions6by their definition tie (2* functions together with a seamless user interface so (2* users can not only share data but can actually run the same (2* modules A marketing person can look at a customer,s contact data through the same application used by the sales force %ikewise, analysts within the company can access a centrali'ed (2* portal to view a range of customer4related data without having to change application tools *icro3trategy,s new e(2* product provides a central portal or BcockpitC through which marketers can create and view customer segments, monitor campaigns in real time, and generate a variety of customer reports "his vendor, long known for its business intelligence software, is combining powerful analysis capabilities to interrelate (2* functions such as Web traffic analysis and marketing automation Database powerhouse /racle continues to enhance its integrated suite of tools, which includes everything from Web site storefront customi'ation to customer self4service, and arch rival 3iebel 3ystems, in addition to offering a range of service, marketing, sales force automation, and analysis capabilities, recently unveiled its 3iebel +andheld organi'ation to further its multimodal capabilities
2,A%
Despite its varied interpretations, Bdata miningC has acAuired an almost mystical allure over the past decade Although its widespread interpretation is as an activity associated with Auerying increasingly detailed data 6Bdrill down,C as it,s called6in fact, data mining is a highly speciali'ed subcategory of analysis that has specific applications from both within and outside (2* !n fact, the term Bdrill downC is more appropriately applied to the practice of online analytical processing, known as /%AP /%AP has become the most popular type of decision4support analysis, allowing the average businessperson to e5plore data online with the aim of focusing on detailed data at a lower and lower level of the data hierarchy *ost often, this means generating an online report, analy'ing the results, and submitting a more detailed Auery in order to understand the result data /%AP generally focuses on providing a set of data attributes from a database organi'ed around certain dimensions, such as time and location "hus, a user can reAuest the company,s regional sales revenues for all baby care products by region or by store +e can reAuest a report detailing regional revenues for each month within a Auarter Although /%AP is generally lumped into the data4mining rubric6usually by software vendors eager to claim the data mining moniker6it normally relies on data that has been summari'ed according to particular dimensions Data mining involves the identification of meaningful patterns and rules from detailed data, usually from large amounts of data "hus, instead of analy'ing customer segments to determine who is likely to churn, as with /%AP, data mining would e5amine individual customers, touching each of the millions of records in a database /%AP analysis reAuires the analyst to have a Auery or hypothesis in mind, but data mining can generate information to show patterns and relationships without the analyst,s knowing about them Data mining can identify clusters of customers who buy similar products6for instance, home office workers who buy P(s, power supplies, toner, printer cables, wastepaper baskets, and coffee With an /%AP tool, the analyst would have to guess which products a home office worker would purchase and then identify customers making such a purchase /%AP analysis typically e5amines category groupings such as P(s, printer cables, and toner <computer4related products?, but might not recogni'e out4of4category purchases such as coffee and wastebaskets
/ne central difference between data mining and other types of decision4support analysis is that data mining usually involves statisticians or product specialists intimate with the use of the correct algorithms and their application to business problems, as well as with the specific data mining software Although a businessperson rarely mines the data herself, she might use data4mining results6either represented graphically in a visuali'ation tool or deployed to a database for general Auery access6to help make important decisions about managing customer relationships "here are myriad uses for the three types of data mining #ust described, from targeting brand new customers by modeling e5isting customers, response patterns to avoiding high4risk prospects through risk prediction or forecasting a customer,s lifetime value *any companies have acAuired dedicated data mining servers, onto which they load customer data records to build models and e5plore various customer behavior patterns 3uch activities are usually processing4intensive so standalone data mining platforms avoid impacting processing on other systems "hese servers are usually linked to a company,s data warehouse, enabling data analysts to easily access customer data to e5periment with various pricing plans, for e5ample, or to create dynamic customer segments for testing new campaigns and performing what4if analysis -ach type of data mining can yield findings that result in high4impact business actions 7or e5ample, an electronics retailer in %ondon discovered that most prospects likely to buy a portable D&D player commute to work on the train, causing the retailer to reallocate much of its marketing budget from daytime television commercials to newspaper ads and billboards "he company saw sales of these players shoot up HI percent after changing its ad media 0nderstanding the impending behaviors of customers and prospects is the key to data mining, and where (2* is concerned, two data4mining applications in particular stand out: click"tream anal&"i" and %er"onalization*
Clickstream Anal!sis
!" departments have become giddy over capturing clickstreams6the data that illustrates a Web visitor,s footprint around the site (lickstreams connote how the user arrived at the site, how long he stayed, what he did during his visit, and when he returned "hey,re the eAuivalent of a camera in a department store recording a shopper,s every move (lickstream data6usually stored either as part of a company,s data warehouse or in a dedicated clickstream data store sometimes called a Bdata webhouseC[4] 6is growing hand4in4hand with corporate e4 commerce activities
[4]
9es, ideally a customer,s clickstream data should be integrated into the enterprise data warehouse along with other customer4related data +owever, much of the clickstream data collected by companies is anonymous data that needs to be mapped back to customer4specific information with the customer,s permission 7urthermore, clickstream data
/ne client of mine, a general merchandise retailer who has #oined the e4tailing ranks, wants its Web site to be as BstickyC as possible and has begun analy'ing clickstream data to surmise why customers might leave the site prematurely "he company has sharpened its analysis to determine the value of abandoned shopping carts When a customer leaves the site in the midst of a shopping trip, whatever the reason, the company looks to see what products were in the cart "he data is then compared with similar data from other abandoned carts to e5amine +ow much revenue the abandoned carts represented <in other words, how much revenue was lost because of the customer,s early departure? Whether the products in the cart were high4profit items or loss leaders !f the same products were found in other abandoned carts "he volume of products and the number of different product categories in the cart Whether the total bill for the abandoned carts consistently fell within a certain dollar range At what point during the shopping trip the cart was actually abandoned <When the customer saw the shipping chargeM When the site reAuired a personal survey before confirming the purchaseM? +ow the average and total bills for abandoned carts compared with BunabandonedC carts6those that made it through the checkout process
"he result of this analysis can trigger some interesting theories 7or instance, perhaps none of the products in the cart was appealing enough to a particular customer to motivate her to continue shopping /r the customer was put off by freAuent inAuiries asking her whether she was ready to check out /r possibly, at a particular dollar total, the customer thought the better of the entire shopping trip and bailed 7inally, perhaps the number or mi5 of products in the cart reminded the customer of another site that offered a steeper discount for similar purchases Admittedly, some of these theories are mere guesses $ut when e5amined regularly and with consistent metrics, clickstreams can reveal some interesting patterns "he fact is, whatever the customer,s reason for leaving the site and a cart full of merchandise, the e4tailer can take a variety of actions based on both hard findings and less4then4certain e5trapolations "he e4tailer can use these results to tweak the design and content of its Web site and monitor resulting improvements Patterns might indicate product affinities, suggesting cross4selling or up4selling strategies And when combined with customer demographics, psychographics, and past behaviors, clickstream data can bring the understanding of customer behavior to a whole new level "he latter option is perhaps the most intriguing: rather than simply e5amining a customer,s navigation patterns and guessing about which actions to take, the retailer can combine those patterns with more specific customer data6his previous purchases in that product category, key demographic and psychographic data, or his lifetime value score, for e5ample6to provide a holistic view of that customer,s value and interests !t might have been a one4time4only shopper who was lost, but in other cases a high4 value customer might have left the site on multiple occasions A tailored e4mail message or electronic coupon6perhaps targeting one of the products left behind on a prior trip6could make all the difference the ne5t time that customer decides to log on "he following scenario, based on a real4life case study, illustrates how clickstream data, when integrated with other key data from around the enterprise, enhances opportunities to personali'e customer communications
+cenario
9ou,re a marketing manager at an up4and4coming eyewear Web site, www glasses4r4us com 9our company has #ust deployed a so4called data webhouse and for the past few days you,ve been looking over your data analyst,s shoulder6they,ve #ust begun gathering user clickstream data 9our data analyst, Jack, points out one session particular 9ou notice that the shopper arrived at your site through a banner ad posted on a partner,s Web site: www eyecare4youcare com After the shopper arrived at your site, he entered his prescription and began looking at various brands of glasses After browsing for glasses, the shopper started comparing brands of contact lenses +e finally selected a brand of contacts and placed the product in his shopping cart +e entered the order information, including his credit card number, but left the site rather than submitting his data and
Along with this more personali'ed tactic, you could also monitor the referring Web site for other referred shoppers who have researched or purchased contact lenses !f contact lens activity is particularly high, you might consider placing a more customi'ed banner ad on the partner,s site and even provide better financial incentives for the partner when new contact lens customers click through With the e4mail strategy, the customer,s more likely to return to the site and you,re almost guaranteed a purchase Win4win, rightM Wrong "he problem with this scenario is that even though analysis is involved, it,s still dangerous "he fact that you,re looking at only a single customer touchpoint can mean big problems and bad decisions !f your clickstream database contained behavior history on this shopper, things might turn out differently 9ou would have more information about the customer, and you,d know the following: "his isn,t the customer,s first visit to your Web site +e has made three other purchases on three separate occasions "he products he has purchased have all been on sale
!n short, you would understand that your Web visitor is what,s known as a Bcherry picker,C someone who only purchases low4margin products when they,re being promoted 1o cross4selling, no up4selling, no true loyalty +e,ll be back again, too, when he finds the ne5t markdown
Personali'ation can take various forms !t can involve customi'ing actual Web pages, including a Web site,s look and feel, according to the features favored by an individual visitor *any Web sites allow the visitor to customi'e the site according to her preferences, eliminating format variations and allowing her a private window into the company 0se the search function oftenM *ove the search window to the top of the page %ike customi'ation, so4called locali'ation can focus site content to the visitor,s particular geographic area 1otice that the personali'ation e5amples from the drugstore e4tailer above hint at predictionM !ndeed, most personali'ation software involves specific data mining algorithms "he two main types of personali'ation are rules4based personali'ation and adaptive personali'ation 2ules4based personali'ation leverages established rules that dictate, for instance, which products might be purchased together or whether a certain Web page should precede or follow another When a visitor to a software Web site buys Ouicken, the site might suggest he buy ?uicken$ The Official @ui)e before going to the checkout screen 2ules4based personali'ation most often involves rules that have been hard4coded into the software 7or this reason, it,s often difficult to maintain and support "he other type of personali'ation, adaptive personali'ation, learns as it goes *ore commonly known as collaborative filtering, this type of personali'ation gets smarter as it observes customer behaviors and applies them to new circumstances 7or instance, if a gardening e4tailer using collaborative filtering observes that shoppers tend to buy low4cost perennial flowers at the same time they order gardening tools, the Web site might begin suggesting a flat of pansies to all customers who buy bulb planters (ollaborative filtering uses the behavior of other BlikeC visitors as the basis for its recommendations (ollaborative filtering tools are often more comple5, and thus more e5pensive, than rules4based personali'ation "he most celebrated e5ample of collaborative filtering is Ama'on com,s purchase circles, in which Ama'on factors in the buyer,s past purchases and geography to suggest what readers who live in her neighborhood and have similar interests might be reading "he more similar shoppers buy, the smarter Ama'on becomes about their preferences, and the more accurate are the site,s recommendations 3everal Ama'on com customers ! know are cherry pickers on other booksellers, Web sites during special promotions, but they always return to Ama'on because Bthey know me better C Perhaps the most telling delineation in personali'ation is in whether or not the user knows it,s happening !n the permission marketing scenario described in Chapter 2, Web visitors voluntarily provide personal information to Web sites where they believe there will be some sort of 6ui) %ro 6uo$ the company will use the information to provide a value4added service such as periodic discounts or special4interest newsletters 3ome sites can personali'e content without making the shopper aware that the products he,s seeing are different from those of fellow shoppers6who might have different profiles and preferences
The Challenges&
+aving begun the process of building a data warehouse back in FGG@, 0nion $ank of 1orway has witnessed its evolution, in terms of both the data and the applications that provide business users with critical information $ecause the bank took a reAuirements4driven approach to (2*, development has proved to be relatively problem4free "he challenge was in building the organi'ation around the data warehouse 1ot having had a centrali'ed location for key corporate data in the past, many of the skills reAuired to develop and support the data sub#ect areas, database, and applications were new to the bank Job roles such as database administrator, application designer, data modeler, and others uniAue to a data4centric organi'ation had to become institutionali'ed to support the bank,s newfound business intelligence environment
)ood Ad9ice&
.ari /pdal is uneAuivocal in her belief that (2* is more of a series of behaviors than it is an actual pro#ect BWe don,t actually use the term internally a lot because there,s so much hype these days,C she admits B(2* can be so easily misunderstood to be something that it,s not C /pdal and her team symbolically point to the FGG@WFGGK period in which the data warehouse was adopted as the hub of (2* /pdal,s advice to others beginning large4scale (2* pro#ects focuses on the clarity of the (2* vision B+ave a clear goal of where you want to go,C she counsels, Band then take it one step at a time C Almost as an afterthought, she adds what might be the best advice of all: B"hink big) act small C
"he integration so widely acclaimed as a (2* imperative has as much to do with systems and business processes as it does with data (ommunicating with customers not only needs to be effective and on4target, it must be seamless (ustomer perception is everything, and customers should perceive that your company applies intelligent service standards regardless of the customer,s preferred channel, desired product, or service reAuest /nly you know how that intelligence translates into initiating the right message to customers 1either 2ome nor an enterprise (2* infrastructure is built in a day !n Part II of the handbook, see how a series of implementation tactics and critical success factors play a critical role in (2* deployment
+cenario
9ou,re a consultant at one of those ubiAuitous Web design and consulting firms 9our company,s stock price has plummeted <unlike staff attrition, which has skyrocketed? *anagement is frantically trying to re4brand the firm from a Web design boutiAue into a bona fide management consulting company A high4profile e4commerce client has recently decided to take on (2* and has come to you for help (het, your company,s retail partner, calls a meeting of the newly formed pro#ect team A debate ensues about what the client means by B(2* C B"hey obviously mean personali'ation,C says "had, a programmer with thick black glasses that label him a hip techno4geek BPersonali'ationMC says (het B"he site,s not even tracking its visitors yet We,ll have to assess their e4business technology infrastructure C BWhat kind of customer data do they have nowMC asks a pro#ect manager as she munches on a muffin B! think they have a marketing data mart,C says another programmer B"hey have to capture clickstreams, we know that much,C comments a database e5pert hovering near the door 9ou,re new and reluctant to weigh in, but your time management ethic overrules your shyness and you ask, BWhat does the client want to do with (2*MC All heads turn in your direction 1obody speaks "he retail partner looks at you as though you,ve #ust fallen off a charm bracelet 9ou sink down in your chair !t,s going to be a long pro#ect (an your company or department answer the Auestion: What !u"ine"" value )o we e-%ect from RMC And if so, is the answer one that will lead to Auantifiable improvements in customer retention and satisfactionM Will it generate profitsM As we,ve seen in Part I of the handbook, (2* isn,t a single product or technology !t,s not e5clusive to marketing or customer care And it ideally involves a cross4section of customer touchpoints !ndeed, (2*,s inherent comple5ity renders it a risky endeavor, even for the most mature companies
B(ritical 3uccess 7actors in Planning, !mplementing and Deploying (2* "echnologies,C working research paper, =>>>, 0niversity of Dayton 8raduate 3chool of $usiness, conducted by 9ancy /shita and sponsored by Dr Jay Prasad, Dept of *!3 and Decision 3ciences
(2*,s ability to impact corporate strategy <according to =@ percent of respondents? Table ?-1. )auging the 0actors o" CR# +uccess Ideal An e5ecutive or board member reads about (2* and understands how its benefits can result in competitive advantage A cross4functional e5ecutive team agrees that (2* is a competitive necessity Desira)le A customer support e5ec returns from an industry conference where a case study depicted uplift in e5isting sales via (2* A business visionary sees Auantifiable benefits for her organi'ation in the short term and for the company at4large soon after "o provide an organi'ation with a greater degree of customer knowledge and improved customer interactions *ndesira)le A product manager sees a vendor demo and returns to the office touting functions and features "he !" organi'ation decides to implement (2* because an e5isting vendor has #ust substantially discounted its (2* software "o automate e5isting processes6 especially if they aren,t costly to begin with /r to add (2* technology to the !" portfolio
3ponsorship
/b#ective definition
!ncreased customer loyalty, better customer service, additional sales revenues, and an overall enhancement of e5ternal perception
/perating environment
!ntegration of (2* product into !ntroduction of dedicated (2* 3tandalone (2* system e5isting !" infrastructure, environment linked to corporate including -2P and data network and key data sources warehouse systems -mployees across the $usiness people from one or corporation at all levels, using two departments leveraging (2* for different purposes but operational and analytical (2* basing their decisions on the same customer information Process efficiencies and integrated data combine to deliver strategic decisions, in turn leading to higher customer profitability, sales uplift, and customer satisfaction Automation leads to process efficiencies and new information that advance departmental goals and result in improved customer satisfaction !mproved perception among e5isting customer base and suspected improvements in marketing campaigns, closed sales, product Auality, and so on /perational (2* available to a select group of users who disperse findings from time to time to selected e5ecutives6on paper Automation leads to process efficiencies resulting in timesavings but failing to cover (2* program e5penses !" has successfully linked the (2* system to operational systems and has deployed (2* to F>> desktops
0ser community
-fficiencies
*easurement (lear sales uplift or decreased complaints and measurable improvement in customer response rates across touchpoints = I H
3uccessful technology integration <=I percent? -nhanced strategic partnerships <=> percent? Assimilation of (2*4related technologies <FE percent?
As Figure 7-1 illustrates, companies participating in the study cited other success factors: end4user desktop workstation configurations, user skill sets, and overall technology architecture 0igure ?-1. CR# +uccess 0actors .Courtes! o" 8ni9ersit! o" 'a!ton )raduate +chool o" Business/
"he point here is that companies implementing (2* understand that the means to the end doesn,t matter6 it,s the program,s ultimate strategic impact and the usefulness of the resulting tool set that affect the perception of success !f the new customer profiling system provides new details about customer behavior but can,t be viewed by call center staff, it,s still a flop And even the happiest campaign managers won,t bolster the dashed hopes of e5ecutives who were e5pecting increased response rates corresponding to (2* deployment
$ecause the company,s customer support organi'ation reAuired basic information about customers and trouble tickets, it was the first department to recogni'e the value of combining operational (2*6the company was surveying customers at the conclusion of every Web4 or telephone4based contact and tracking customer satisfaction scores6with analytical (2* to streamline its call center processes "he call center,s goal was to use survey scores to analy'e customer complaints and foster product and service improvements while simultaneously putting in place a Web4based customer self4service infrastructure /ver time, the company,s customers would be able to reAuest service on the company,s Web site, mitigating the need for in4person assistance, in addition to being able to order new services and add4on features such as caller4id (ustomer support recogni'ed the promise of not only cost reduction but also higher customer retention rates
?uantit& of function"* !f your (2* ob#ective is simply to deliver customer profiling, you probably have a single function !f it,s to automate your campaign management, you,ll likely have at least a handful of functions to implement <We,ll discuss functionality in more detail in Chapter 8 ? Range of u"age* +ow many departments are slated to use the (2* system after it,s up and runningM !mplementing (2* for a single relatively small department is much less comple5 than deploying it to the entire enterprise
"he contrast among the four Auadrants in Figure 7-6 is stark and has significant impact on the development process, as illustrated here: A single4function (2* pro#ect to one department is nothing more than a customer4focused application !t is most likely driven by a handful of business people and managers, not corporate e5ecutives, and will be used by a single organi'ation 9ou,ll probably be able to leverage a series of in4house development processes and e5isting staff to deliver single4function (2* to the department that needs it A multifunction (2* pro#ect to a single department is another story !nstituting a customer4focused contact center dictates a range of new customer4oriented business processes, not to mention new policies and end4user training Defining and documenting business processes, as we,ll show in this chapter, will give you a good idea of the (2* system,s true comple5ity and the development resources it reAuires (onversely, a single (2* function to be deployed across the company represents a newly institutionali'ed business function !f the call center, marketing, risk management, and sales
"he differences among the four Auadrants can influence everything from e5ecutive level involvement6 unnecessary for single functionTsingle department, mandatory for multifunctionTenterprise4wide6to the range of reAuired technologies, development skills, and end4user involvement !ndeed, the top right Auadrant of Figure 7-5 points to the role of (2* as not #ust an application or pro#ect, but as a cor%orate a""et to be deployed and managed on behalf of the company "he nature of this approach is both information4centric and customer4focused !t suggests treating customers themselves as a corporate asset, given the same amount of money, infrastructure, and e5ecutive attention as other corporate assets, if not more While comple5ity should be the key metric in estimating the cost, resources, and development steps necessary to implement (2*, your company,s si'e is also a factor A large company, for instance, might have the skills and infrastructure to dedicate to a (2* pro#ect and is probably adept at handling large4scale enterprise systems development 3mall to mid4si'e companies, on the other hand, won,t have as many organi'ations or the same number of stakeholders as larger firms "hough this can streamline business planning and vendor selection6fewer players simplifies consensus4building6securing funding might be more challenging for a smaller firm 1ailing down the right e5ecutive sponsor for (2* is probably also more straightforward in smaller companies
-ven if your company doesn,t have a structured program4approval process, including a discussion of each of the above considerations in your (2* business plan will ensure you,ve done your research and help bolster your arguments
!ts adherence to +ow (2* pertains to the company company,s stated goals or ob#ectives overarching strategies !ts ability to deliver key business ob#ectives !ts cost
+ow specific business goals (2* will allow the company to adopt true one4to4one will be met with (2* relationships with our customers by delivering both personali'ation on our Web site and real4time customer profiling capabilities for our call center staff An estimate of the cost breakdown During the ne5t fiscalyear, the proposed (2* program is estimated to need DF @ million in technology funding <hardware, software, networking?, an additional DF million for permanent head count, DF @ million for consulting services, and a half million for e5ternal data acAuisition "he initial release of the e(2* program will include deployment of !&2 self4service, Web4enabled provisioning, and Web 7AO services to alleviate demands on the contact center !n addition to the current (2* 3WA" team, we estimate the need for A (2* development manager <7"? "wo (2* product specialists <7"? A (2* architect<consultant? An additional database administrator <7"?
!ts boundaries
An e5planation of the initial (2* pro#ect,s resulting deliverable A list of necessary staff for reAuirements gathering, technology acAuisition, development, and rollout of the (2* solution
3taffing reAuirements
2isk assessment A description of the potential We foresee the e4business organi'ation,s historical risks involved in launching a reluctance to share its data as a likely impediment to sales4 (2* program at this time department access to e5isting customers, Web purchase and self4service history, rendering customer history profiles incomplete and the resulting decisions potentially faulty (onsider the following two lists of (2* business Auestions taken from actual pro#ects "he first list, from the marketing department of a cable "& company, represents a list of departmental reAuirements: Product managers must be able to define their own campaigns
"he ne5t list also originated from a marketing department, this time from a large communications company +owever, this list addresses cross4functional reAuirements: /ur campaign managers need to know if their recent campaigns have resulted in increased customer support reAuests /ur product planners are interested in whether e5isting product usage rates affect new campaigns We,d like to analy'e whether a campaign will be more successful with customers who already spend DF>> per month .nowledge of whether direct sales and reseller channels influence campaign success can help both sales and marketing optimi'e channels and direct sales staff We need to know which campaigns were more effective with resellers With the Web and e4mail marketing With direct sales <telemarketing? 3ales management would like to know if campaign effectiveness is related to the length of the reseller,s relationship with us 3ales management wants to relate the success of a given campaign with sales compensation and commission levels
1otice how the reAuirements in both lists pertain to marketing improvements, but the items in the first list are e5clusive to marketing and can thus be considered departmental "he cable company,s marketing department has identified the areas in which it can improve campaign effectiveness and optimi'e customer interactions, both worthy (2* ob#ectives (ampaign effectiveness is important in the second list, as well) however, the communications company,s reAuirements involve both a cross4section of users and a greater variety of data "he first reAuirement, for e5ample, involves analy'ing customer4support trouble tickets before launching a campaign, and the ne5t several involve data from other systems, such as the provisioning and billing systems "he last several reAuirements describe how (2* will aid the sales organi'ation 9our (2* business plan should not only list such customer4focused reAuirements, but should also map them to specific applicable (2* tactics, providing management with a reconciliation of which (2* features will address which business goals <And as we,ll see in the ne5t chapter, this also renders technology selection a whole lot easier ? Table 7-3 illustrates the mapping between a set of cross4functional business reAuirements and specific (2* tactics "he advantage of this type of matri5 is that it provides a visual clue to critical (2* capabilities6notice how prominently personali'ation plays a role with most of the business ob#ectives6while also presenting a good idea of what will be involved in reali'ing the business reAuirements 7or instance, the Bincreased service and repair effectivenessC reAuirement ultimately warrants a series of non4(2* features to be successfully implemented, meaning (2* in field services might reAuire more resources and take a bit longer to deliver "o ma5imi'e the success of the first proposed (2* program, the program needs business reAuirements that +ave defined boundaries +ave a high value4to4cost ratio
!mprove work efficiencies for more than one person !nvolve process change
!t is the cost4to4value ratio that most confounds well4meaning managers who intuitively know (2* is the best weapon, but who need some ammunition
Cost-Custi"!ing CR#
When launching a visible and wide4ranging program such as (2*, it,s only a matter of time before a high4 ranking e5ecutive inAuires, B3o how much money have we spent on this (2* thing, and what have we gotten in returnMC "he degree to which your (2* program has been deliberately planned and e5ecuted is the degree to which you,ll have a slam4dunk answer to this Auestion Any (2* program has three possible financial outcomes: F = I !ncreased profits $reak4even %ost revenue
0nfortunately, Auantifying how much additional profit is generated or money saved via (2* is difficult 0nlike more straightforward operational systems that deliver both defined outcomes and Auantifiable improvements, (2* often fosters unprecedented business practices that are by their very nature not measurable (omparing new sales channels such as the Web to traditional channels invites apples4to4 oranges debates 7urthermore, unlike its more technology4specific counterparts, (2* often delivers 2/! that is both hard and soft 7rom a soft return standpoint, (2* can deliver significant payback that,s nevertheless difficult to Auantify -nhanced employee satisfaction, cultural and workplace improvements, perceived technology leadership, and amplified market reputation are e5amples -ven such concepts as customer loyalty and customer satisfaction, both crucial to business success, are difficult to measure A *arch =>>> (2* study conducted by *-"A 8roupT!*" revealed that G> percent of the fifty largest (2* user companies admitted being unable to Auantify a return on their (2* initiatives 7or some companies, simply knowing that, after deploying (2*, their sales figures e5ceeded the industry average is enough 7or others, the inevitable e5ecutive Auestions loom large6large enough to mandate tangible benefits
"ake the first item on the list as an e5ample "he Director of Product *arketing at a large regional bank described it in practice: 2ight now the bank can,t keep track of more than F> campaigns at a time We want to create product offerings that are uniAue to specific customer segments, which could increase the number of campaigns ten4 or even twenty4fold We desperately need to manage more campaigns in order to promote more distinct offerings 2eally, we,d like to move toward one4to4one, where instead of having one campaign for a million consumers, we have a hundred campaigns each focused on a group of F>,>>> consumers "his strategy will increase response rates for our marketing campaigns and generate additional revenues !n fact, every business ob#ective you define as part of your long4term (2* planning should inherently target one of the three metrics 7or e5ample, we could map each of the ob#ectives listed in Table 7-3 to one or more of these metrics, as we have done in Table 7-4 <where, again, BAC signifies business process efficiencies, B$C decreased customer attrition, and B(C increased sales? "he most straightforward way to calculate the financial promise of one of these business ob#ectives is to measure how it,s currently being done and what it,s costing 0nfortunately, because many of these initiatives involve new corporate paradigms, there is often nothing to measure /ne client of mine understood the degree to which its call center staff was spending time on unnecessary work "he company had hired a consulting firm to measure (32 activities and determined that (32s spent an average of =@ minutes on every trouble ticket simply gathering customer data "he company then determined that, of these =@ minutes, F@ were spent accessing and logging onto various source systems, searching for specific data, and consolidating the answer sets Table ?-;. Business 2bGecti9es and 0inancial #etrics F 8reater number of Web site return visitors = -4commerce efficiencies I !ncrease in market share for core products H +igher customer4satisfaction ratings @ !ncreased campaign response K !ncreased service and repair effectiveness A, ( A, ( ( $ A, $, ( A, $
We used the following form as a way of measuring this company,s (2* opportunity: %roblem +tatement& /ur call center staff productivity has decreased dramatically as the problems become more comple5 We need a means of increasing (32 productivity to improve the cost structure of the call center +ample problem -uanti"ication& -veryone knows the number of trouble tickets e5ceeds the e5isting staff,s ability to process them We,ve recently determined that the average (32 can handle F> tickets a day "he average amount of time spent in data4gathering <which includes accessing data from five different systems? is =@ minutes per ticket BA single (32 tool and screen should be able to reduce data4gathering time and allow our (32s to address more trouble tickets in a given day C <&ice President of (ustomer 3upport? 1umber of (32s X K> Average time to gather customer information X =@ minutes 1umber of tickets generated for each (32 per day X I>WH@ 0iscal premise& Average yearly burdened cost of (32 X DK>,>>> Average tickets per (32 per day X F> (ost per ticket X D=@
Impro9ement -uote&
2perational premise&
2educed trouble4ticket response times !mproved customer4satisfaction levels !mproved employee4satisfaction levels
!n a true enterprise (2* opportunity, several such forms are completed and then compared "he highest4 impact (2* opportunity inevitably rises to the top, becoming the first official (2* pro#ect within the (2* program +ere,s another (2* measurement form completed for the same company, this time for the marketing department: !n the event that several of these forms can be completed for each of your company,s (2* applications, they can be collected, prioriti'ed, and then published in a physical or online document to serve as a living (2* roadmap %roblem "he company,s marketing process is too darn long +tatement& +ample problem -uanti"ication& Impro9ement -uote& 2perational premise& 0iscal premise& !t takes up to K weeks to identify a campaign target audience6using e5perienced data analysts We,d like this to take days, or even hours, using marketing staff with minimal assistance from !" B$y reducing the time needed to identify a campaign,s target audience, we could double or even triple the number of campaigns we deploy, while further delimiting our target segments C <Director of 3egment *arketing? 2ight now, for every three marketing campaign managers, we need one data analyst and one !" Auery support staff member to run Aueries -ach campaign manager reAuires two support staff members: Average yearly cost of F data analyst X DFI>,>>> Average yearly cost of F !" resource X DFI>,>>> 1umber of campaign managers in marketing X I> CR# impro9ement assumptions& (ampaign managers will migrate to using desktop (2* analysis and will need to be educated on its use (ampaign managers will evolve from pro#ect managers to Bknowledge workersC "he pro#ected cost savings will occur via the reduction of data analysis and !" support staff Productivity gains will increase the number and effectiveness of campaigns by a minimum of => percent Y of support staff K FK =K Pro#ected staff savings<nZ DF@>.? DG>>,>>> D=,H>>,>>> DI,G>>,>>>
A good e5ample of (2* 2/! is 3$( (ommunications According to IO maga'ine, 3$(,s -A3- <-asy Access 3ales -nvironment?, an online tool to help (32s and telemarketers, cost DIH = million "he tool helped sales reps access product information faster and pull in more comprehensive customer data, allowing an increase in call volume, a simultaneous decrease in call duration, and improved order accuracy All told, the estimated 2/! for -A3- was reported to be DHEI K million [2]
[2]
+ere,s an important note on 2/! financials if you,ll be asked to #ustify (2* to a (hief 7inancial /fficer or corporate accountant: ! had the pleasure of e5plaining (2* 2/! to a finance director who asked if (2* was worth more than the Btime value of money C "his manager,s Auestion implied that his company would either invest in (2* or put the money into some interest4bearing checking account6which wasn,t the case !n fact, cost4#ustifying a (2* program for most companies assumes that the allocated budget money will go either to (2* or to another proposed pro#ect of similar or higher priority "he goal is to make a case for the value of (2* versus other potential corporate programs Although e5amining the time value of money might be worthwhile if your choice is either to implement (2* or to purchase real estate for a new brick4and4 mortar store, the ma#ority of companies considering (2* are going to spend their money on a pro#ect "he issue is where to get the most bang for their budget buck Another cost to consider when #ustifying (2* is the cost of delaying the decision to move forward 7or instance, in one comprehensive (2* business case, one client of mine included a section citing the following opportunity costs of delaying (2*: "he cost of lost marketing opportunities, including o o o (ost of lost customers due to competitive marketing events 2educed effectiveness of new products due to lack of market understanding (ontinued increase of marketing costs due to poorly focused campaigns andTor oversi'ed target audiences
(osts of continuing the support of stovepipe database systems %oss of staff skills and e5perience due to staff redeployment %ost !" resource and sub#ect4matter e5pertise due to normal staff attrition rates 2educed customer loyalty and perception due to inability to enhance the customer,s relationship e5perience
!n fact, many companies that undertake (2* are measuring their successes based not on return on investment but on return on relation"hi% <2/2? 2eturn on relationship implies the ability to compare the before4and4after effects of (2* on customer value and loyalty +ave customers in the mid4tier sector migrated upward in valueM +ave customers we,ve been Bwilling to loseC actually become more profitable since (2* was establishedM +as a low4value customer referred high4value customers to the company, thus contributing even more revenue than if she had spent that money herselfM *easuring 2/2 can be sub#ective but can provide a company with the ability to identify which components of (2*6be they changes in business processes or more targeted communications6improve customer relationships and which customers seem most responsive to new customer4focused business actions "hen the company can formali'e what,s working and fine4tune what,s not A final word on (2* 2/!: !f your company is adopting (2* because of the cost savings it promises, beware With the escalating comple5ity and pricing of many (2* products, seeing return on investment might take a few years "he real #ustification for (2* goes back to improving your customers, e5perience with your company, humani'ing this e5perience, and making it easier to do business with you (2* is about managing and monitoring your customer relationships and increasing their value !t,s about motivating customers to tell their friends to buy your products 9es, these too can result in revenue down the line $ut when a good customer is #ust a mouse4click away, delaying (2* can be risky
"he good news was that the hotel company was already differentiating its high4value customers6one of the main tenets of (2* 7reAuent4guest card4holders had their own toll4free number directing them to call4 center agents trained to take their time when dealing with high4value guests "he bad news was that if the customer needed the room right away, he was normally unwilling to risk being wait4listed and would simply call a competing hotel chain "he company,s data warehouse verified that of all freAuent guests being refused reservations on their first attempt, only =H percent agreed to be placed on a wait4list, and a mere FF percent actually ended up staying at one of the chain,s properties
"he resulting freAuent4guest reservations policies dictated a more speciali'ed reservations process that looked like the one illustrated in Figure 7-8 0igure ?-@. The new customer-"ocused business process
1otice that the new process has not only been rendered more involved) much of it has been automated, making it easier for the customer to communicate with the company "he rule of modeling customer interactions is that every interaction, incoming or outgoing, should have the potential to improve the customer,s e5perience $y offering its best customers a choice of media for making reservations, the hotel effectively provided more value to the customer "he BagentC referred to in the e5ample might be either a human (32 in customer support or a cyberagent interacting with the hotel chain,s reservation system as well as those of its partner hotels <who reimburse the hotel chain with a referral fee? As we discussed in Chapter 2, the use of cyberagents to automate decision4making and accelerate previously manual processes is another effective way to speed up workflow !n fact, the term BworkflowC is used in (2* to refer to automated business processes *any (2* products feature Bworkflow managementC components automating processes such as campaign management or customer troubleshooting "his hotel chain knew it needed to implement both operational and analytical (2* "he chain,s customer support department understood the business need, and even knew where it wanted to begin (ustomer support specialists and business analysts mapped out the new reservations process, focusing on the various customer touchpoints to ensure that individual interactions represented an improvement over the traditional process $ut in this case the whole was greater than the sum of its parts !t was a BsoftC benefit of (2* that ultimately delivered the biggest payback: the hotel chain was also able to increase its brand loyalty +igh4value customers gradually reali'ed that when they called the chain,s freAuent4guest reservations line, their likelihood of getting a room, even if not in a first4choice property, was higher than before6and much higher than when calling a competing chain (ustomers no longer had to waste time calling around to different
!f you don,t know the answer to two or more of these Auestions, you would do well to take the time to map out new or e5isting processes and identify areas that can deliver an improved customer e5perience and tighter time frames !n addition, try looking at your business processes from an organi'ational perspective *ost process planning activities neglect this step, but answers to Auestions like those in the following list can result in even more highly refined processes and can pinpoint opportunities to improve your overall infrastructure: 7or a given customer4facing business process, how many departments are involvedM +ow many actual staff members touch in each processM What data is transferred between organi'ations, and how muchM Does the information being shared change as it goes through the processM +ow oftenM Do the organi'ations involved in each business process agree on business rules and common terminologyM
When designing and documenting new business processes, it,s helpful to understand not only the customer,s view of the process but its inherent comple5ity A customer,s potential delight at a new, Web4 based order process won,t matter much if the process itself is too cumbersome to program and integrate with e5isting systems *any process4design teams get caught up in modeling conventions and documentation tools !f you have such tools in4house and the e5pertise is handy, having an online process library enables the company to maintain a history of customer4focused improvements over time 3uch a library can be part of a corporate4 wide knowledge4management system and can be used for a variety of purposes $usiness analysts can access outdated processes to provide (2* stakeholders with graphical illustrations of before4and4after processes during reAuirements gathering Developers and programmers can use the models to ensure that the (2* system mirrors the process vision "he main goal of business process modeling for (2*, however, is to improve traditional or broken processes and thus enhance customer interactions !f you need to initially document processes on a whiteboard rather than waiting to install the latest graphical modeling tool, do so "he convention is not as
After the data warehouse was up and running, the company could periodically supplement its data with various data sub#ect areas from both internal and e5ternal data sources, at each stage delivering enhanced (2* functionality
The Challenges&
%eonard,s challenges were more cultural and organi'ational than tactical 3he cites the need for gathering consensus around (2*,s value as a ma#or hurdle6one she,s gratified to have overcome 3he also charges various product vendors and consultants with glibly promising unrealistic BAuick winsC at the e5pense of the planning and rigor ultimately crucial to instilling the corporate consciousness of (2* as a true enterprise program B(2*, if done right, is not simple,C %eonard declares /ther challenges, such as harnessing individual proclivities for building single4purpose, single4user systems, are ongoing *oreover, the comple5ity of erecting enterprise (2* in a high4profile public company means unremitting pressure from the financial community with its push for immediate 2/!
)ood Ad9ice&
%eonard emphasi'es that amidst the cumbersome analyst reports and comple5 systems comparisons, her team remained mindful of the customer,s perception, as well as her company,s strategic direction B0nderstanding how we want to be perceived as a company has ma#or implications on how we approach (2*,C she says BWe built our (2* model to emphasi'e our brand and reinforce our image C !ndeed, the linkage between the company,s (2* direction and its brand is a top priority 7ar from hopping onto the crowded (2* bandwagon because of its current popularity, &eri'on (ommunications has embraced enterprise (2* for a series of captivating and overarching strategic reasons At &eri'on, (2*4thinking begins strategically and then subseAuently drives planning and development %eonard suggests obtaining solid answers to the following enduring Auestions:
B! want a portfolio of tactics that ! can mi5 and match depending on the customer, event, situation, cost versus benefit, and availability of internal resources6as in the number of service reps who might be available to handle a new product rollout, C %eonard insists Although it supports the company,s brand image, the (2* mantra at &eri'on is refreshingly tactical: B!ntegration, reusability, and cross4functionality C Where should a company startM %eonard,s background in strategic planning has served her well, and she insists that building a (2* roadmap is an indispensable step and internal communication is crucial When positioning (2* to her management, %eonard admits, B! told them it wasn,t Auick, it wasn,t easy, and it wasn,t cheap ! told them they couldn,t go out and #ust ,buy (2* ,C
J (ommunicated strategic initiatives can be !f the company has a list of strategic ob#ectives, those supported by (2* ob#ectives should be customer focused and thus supported by (2* E 3takeholders can articulate pro#ected (2* benefits for each discrete opportunity $usiness sponsors or management should be able to describe the tactical business improvements that can be delivered by (2*
G 3tated opportunities can be improved with "he (2* opportunities being discussed must be able to be customer4related data supported andTor improved with clear, consolidated customer data <!n other words, process improvements aren,t enough ? F> Pro#ected data sources are highly regarded for data accuracy and integrity Where will the customer profiles and segments originateM !f those systems aren,t trustworthy, no one will trust the ultimate (2* applications
FF (ross4functional customer data e5ists in a A data warehouse containing consolidated customer data warehouse or centrali'ed database information from around the company will #ump4start any (2* program and will decrease the infrastructure costs F= /rgani'ations currently share a cross4 section of information reAuirements FI "he client is already engaging in some sort of customer differentiation or segmentation FH Ouestions of data ownership across the company are either none5istent or easily resolved +as data sharing been institutionali'ed already with other systemsM "his is a positive sign, particularly if the initial (2* pro#ect evolves toward enterprise (2* !f customer segments are already being identified, there is an understanding of customer differentiation, which makes (2* much more culturally palatable !n addition, certain e5isting segmentation or analysis process might be leveragable Are specific organi'ations willing to share their data with the rest of the companyM !s management willing to enforce thisM *issing pieces of the customer pu''le could #eopardi'e an entire (2* program
F@ $usiness units and !" staff agree on (2* "he e5tent to which one organi'ation wants to BownC (2* is ownership boundaries the e5tent to which politics will get in the way of productivity "here should be firm boundaries for who does what FK -5ecutive management has an e5pressed -5ecutives should understand not only that (2* involves a commitment to fund (2*4related significant invest4 ment, but that additional funding dollars activities might also need to be reserved FJ (lient agrees to modify business processes as a result of (2* Access to complete customer data should trigger business efficiencies
FE "here is willingness to sustain the *anagement should be aware that, along with more data and organi'ational impact of (2*<for e5ample, process changes, #ob roles might change and new skills might reorgani'ation or additional staffing? be needed FG A general understanding of reAuirements4 3uccessful (2* pro#ects are Btop down,C meaning that they driven development e5ists among both are driven by business need /nce understood, business business and !" stakeholders reAuirements and their relative impact should drive (2* implementation priorities => *anagement is willing to empower key !f employees such as salespeople and (32s have more customer4 facing staff based on increased information, it follows that they can be more self4directed information and improved processes Accountability should be maintained as employees are given more freedom, the focus being on ultimate improvements in customer satisfaction and revenues =F *anagement is willing to implement incentives or modify employee compensation to encourage (2* adoption == 1o decisions have been made about 3taff members who readily adopt (2* technologies and processes, and who participate in their ongoing improvement, should be rewarded 3taff members who refuse to adopt these improvements can be considered Bsaboteurs C Penali'e them $eware the tail that wags the dog: are stakeholders
=I $usiness sponsors and stakeholders have an understanding of the differences between (2* and other programs<such as business intelligence, -2P, or data warehousing? =H !" staffing infrastructure is in place to support (2*
=@ "here is consensus that (2* is a process %ike other large corporate initiatives, (2* is an ongoing and not a one4time4only activity process that grows and improves over time =K $usiness and !" stake4holders understand $ecause (2* is a process, it reAuires ongoing budget that (2* reAuires ongoing budget to $eware the lump sum (2* allocationPit probably won,t cover support continued development and all necessary (2* functionality maintenance Part of readiness assessment involves weighting to specific factors in the evaluation based on the results of the interviews conducted <7or e5ample, if upper management is advocating an enterprise (2* initiative, the e5istence of cross4functional customer data would receive a higher weighting ? 2egardless of weighting, you can gauge your (2* readiness with the following rating scale: H: "his statement is 9er! descripti9e of our environment I: "his statement is largel! descripti9e of our environment =: "his statement is partiall! descripti9e of our environment F: "his statement is not at all descripti9e of our environment
7or e5ample, Table 7-6 shows how a specialty retail client scored on the assessment, and how to interpret its score Table ?->. CR# Readiness +coring F "argeted business users display an understanding of (2* and accompanying benefits = *anagement displays an understanding of (2* and accompanying benefits I (2* application opportunities are identifiable H A business sponsor e5ists for each discrete (2* opportunity F7IH F=:H F=:H F7IH
@ /bvious stakeholdership <sub#ect matter e5pertise, targeted end4users? e5ists for each discrete F = : H (2* opportunity K (lient has e5pressed a need for market differentiation <or similar strategic ob#ective? J (ommunicated strategic initiatives can be supported by (2* E 3takeholders can articulate pro#ected (2* benefits for each discrete opportunity G 3tated opportunities can be improved with customer4related data F> Pro#ected data sources are highly regarded for data accuracy and integrity FF (ross4functional customer data e5ists in a data warehouse or centrali'ed database F= /rgani'ations currently share a cross4section of information reAuirements FI "he client is already engaging in some sort of customer differentiation or segmentation FH Ouestions of data ownership across the company are either none5istent or easily resolved F@ $usiness units and !" staff agree on (2* ownership boundaries FK -5ecutive management has an e5pressed commitment to fund (2*4related activities FJ (lient agrees to modify business processes as a result of (2* F=I; F=I; F=:H F=:H 1=IH F=:H F=I; F7IH 1=IH F=:H F7IH F=:H
FE "here is willingness to sustain the organi'ational impact of (2* <for e5ample, reorgani'ation or F = I ; additional staffing? FG A general understanding of reAuirements4driven development e5ists among both business and !" stakeholders => *anagement is willing to empower key customer4facing staff based on increased information F=:H F=:H
=K $usiness and !" stakeholders understand that (2* reAuires ongoing budget to support ongoing F = I ; development and maintenance "otal 2eadiness 3core: 9ou can now interpret the results, using the following scoring metrics: 1H;I@<& 3uggests your organi'ation is ready to begin implementing a (2* pro#ect with minimal infrastructure enhancement and confidence of a high degree of sponsorship @;I?:& 3uggests your organi'ation should solidify its infrastructure, skill sets, and e5pectations but should e5pect to launch a (2* pro#ect in the near future Planning should begin for a proof4of4 concept ?7I<H& 3uggests your organi'ation should refrain from embarking on (2* until the technology infrastructure, data ownership, or cultural and political issues are resolved 3ponsorship should be cemented and staffing enhanced at this time ;A or below& 9our organi'ation has not e5pressed a firm business #ustification for (2*, or must perform a ma#or overhaul of its staffing andTor systems Another readiness assessment should occur after the identified improvements have been made ?>
+cenario
9ou,re an e5ecutive at a large multinational technology company 9ou,ve recently begun reading more about the use of customer relationship management to instill customer loyalty and discourage customers from doing business with your competitors And you,ve been hearing more about (2* than ever6your &P of *arketing mentions it in practically every e5ecutive staff briefing /ne day she strides into your office and takes a seat B!,ve #ust reviewed a (2* tool, and we need to get it,C she says preemptively as (heryl, your assistant, hovers at your doorway in a vain attempt to foil yet another unscheduled appointment BWhat was the productMC you ask "he *arketing &P states the name of a familiar4sounding company you suspect is owned by one of your mutual funds and begins rattling off bu''words as though she,s #ust learned a new language 9ou don,t know what a screen pop is, nor do you understand what she means by a Bcustomer knowledgebase,C but you let her finish When she does, you ask what any self4 respecting e5ecutive beholden to a bunch of impatient stockholders would ask: B+ow muchMC B"hree million,C she says As if it would e5plain everything, she adds, BA million for the base (2* platform, and another two to re4do our campaign processes and customi'e the code C 9ou,re still wondering why an innocent4sounding screen pop can cost three million dollars "here must be cheaper screen pops to be had B3houldn,t we look into some alternativesMC you ask B*aybe there,s a lower4cost package that does the same thingP C BWe can;t wait that long>C cries the &P B/ur customers are only a click away from one of our competitors: We have to start integrating our channels now: !t costs si5 times as much to get a new customer as it doesP C 9ou,ve seen it hundreds of times in your long career as a manager, and few things are as dangerous and as fearsome to behold: a businessperson who,s #ust returned from a trade show
"hese reasons range from acceptable to dangerous After all, who could argue with impressed users willing to fund development <even though the tool,s functionality might not meet these users, needs?M And it,s certain that, after they,ve approved the (2* business plan, e5ecutives will want to know when you,ll be finished and three4month delivery could make you a hero As for competitive pressures, it,s a worthy e5cuse, but are you sure you know what your competitor means by (2*M Allowing technology to drive (2* is known as the Bbottom4upC approach 0sually this involves one organi'ation, or even an individual manager, who decides to go it alone, allowing a (2* software tool or a specific functional goal to define the (2* deliverable "he #ustification for bottom4up (2* is usually the time reAuired by more rigorous, reAuirements4driven planning "he risks inherent in bottom4up development, however, are far more serious than the rewards: %imited consensus about (2* goals risks spending of money on low4priority capabilities 3ub#ective interpretation of the importance of the given functionality invites rework and wasted resources %ack of integration with other technologies or (2* pro#ects results in either throwaway work or cumbersome after4the4fact integration Dependence on specific product features, which might or might not meet additional business or growth needs, #eopardi'es broader (2* adoption and growth
Although it might be tempting to espouse the well4worn aphorism B!f you build it, they will come,C the truth is that if you build it, they probably won,t even notice After all, how much of your company,s software products have ended up as shelfwareM $ottom4up (2* development means (2* in a vacuum, a pro#ect not reAuested by or sociali'ed to the business 2eAuirements4driven (2*, on the other hand, establishes a level of cross4functional consensus in which people participate from the beginning of the (2* program and thus feel they have a say "he fact is, when choosing your (2* technology, there,s simply no substitute for allowing structured reAuirements to dictate your technology decision 9es, it takes longer than the knee4#erk development with the tool du #our $ut the alternative is much riskier, and e5amples abound of (2* systems that never delivered the goods
A customer4focused business strategy drives a series of (2* reAuirements <e g , B"he ability to track success of target marketing campaignsC? "hese reAuirements in turn elicit specific functional capabilities <e g , Bcampaign response modelingC? And, as we,ll see, when the functionality is understood, a list of products can be mapped to each specific function
[1]
1ote that we aren,t launching a discrete campaign here, but rather simply providing the scoring information to the call center to refine customer telemarketing suggestions !n other words, as telemarketers use the new information in their customer conversations, marketing collects the data and validates that it has helped improve product sales
1otice in Figure 8-2 that the functionality dictated by the business reAuirement is also dictated by the envisioned marketing process !n other words, BAnaly'e customer purchase historyC is a function that supports the reAuirement) it,s also a step in the campaign process "his is the ideal way to define necessary functionality, ensuring that it not only maps back to the business reAuirement but is involved in a business process as well 0igure @-7. Re-uirements J process K 0unctionalit!
"here are many types of campaigns, and the product recommendation scenario described earlier is one of many e5amples "he point here is to define your own uniAue reAuirements After you,ve listed your business reAuirements, as described in Chapter 7, record the functionality for each one, asking, B+ow will we accomplish this reAuirementMC 9ou,ll eventually have a list of key functions for each reAuirement, and you,ll have finished the hard part
When you begin assembling a short list of (2* vendors, a good idea is to get your !" department involved if you haven,t already 1ot only will !" staff understand product technical features and standards, but they,ll also be able to offer guidance on how a given product fits into your company,s systems and data management environments And because !" provides corporate4level support, !" staff might have insight into whether products with similar capabilities have already been evaluated by another organi'ation within the company, or whether other pro#ects need similar (2* capabilities After you have finished narrowing down your product choices and have arrived at a short list of vendors, you should seriously consider formali'ing !",s involvement in the pro#ect 0p until now, you might have chosen to involve key !" players, or even the (!/ in (2* strategy development or planning activities !",s participation in vendor technology discussions is now imperative !f your choice comes down to two vendors, the one that most closely aligns with your e5isting technology infrastructure should be declared the winner
'roce""ing an) %erformance re6uirement"* !ndicate the product,s ability to support and control reAuired operations: o o o o o 1umber of transactions the product can support <upper limit? Data volumes the product can support 1umber of concurrent users that can be supported 3upport of data and system backups 3ynchroni'ation strategy and effort reAuired
Ability to push pre4formatted reports to end users Ability for end users to perform ad hoc reporting Allowing for end4user creation of local <client4based? canned reports Ability for end users to e5tract data for local analysis
9"a!ilit& re6uirement"* -nabling end users to easily and intuitively accomplish reAuired tasks: o o o o o Ability for end users to have a custom home page <Bmy(2*C? Ability for end users to seamlessly access other corporate systems through a common (2* portal Ability to customi'e online help screens with application4specific information Ability to perform screen4prints Ability to display graphics, pictures, and photographs
(unction#ena!ling feature"* "he way in which the product provides certain reAuired functionality: o o o o A workflow management capability, including the support and automation of user4defined workflows An e4mail response engine able to route incoming e4mails to specific (32s Predictive modeling functionality <e g , to apply a customer,s propensity to buy to list4 generation activities? 3upport of wireless access to (2* server
'erformance re6uirement"* %aying out acceptable turnaround time for (2* activities or reporting response time: o o Ability to provide I>4second or less reporting4response time for !nternet users Ability to generate campaign lists, irrespective of attributes, within one hour or less
Availa!ilit& re6uirement"* !ndicating the acceptable level of system availability, for e5ample: o o o o Product and accompanying database both available from E a m to E p m seven days a week !nclusion of self4diagnostic tools that can alert system administrators to slow response or likely downtime Accommodation of different time 'ones if your company is geographically dispersed Web page availability =H4by4J
After you,ve covered the necessary bases in terms of both functional and technical reAuirements, you,ll be ready to have a substantive conversation with your prospective (2* vendors, not #ust about how great their tools are, but also about how well they correspond with your uniAue needs Talking to CR# =endors
Depending on the comple5ity of the software product and it implementation reAuirements, a vendor might be unwilling because of the consulting costs incurred 3ome vendors will be willing to offer a proof4of4concept that allows a prospective customer to try out the software with a limited scope 3uch a proof4of4concept has a price tag covering the vendor,s costs, but the charges can then be applied to the product,s purchase price if the proof4of4concept is successful
Table @-1. Buestions to Ask Dour CR# =endor 'he Domain &endor -5pertise 'he +uestion to As What ma#or (2* functions does your product suite provideM ,hy -ou Should As It "he vendor should be able to provide a good description of every product or module and its relative functionality !t should be clear whether the product represents a (2* suite that provides a range of functions or whether the product is a (2* point solution
!n cases where you don,t offer "he vendor,s strategic alliances can be important clues certain functionality, do you partner about areas where the product might be weak !f you with other companiesTproductsM pro#ect needing additional functions not offered by the vendor, make sure you understand the details of its alliances with other (2* vendors (an you describe how your product Although it probably won,t change things, has evolvedM understanding where the (2* vendor got its start is always helpful "his doesn,t mean the company that started as a call center software tool can,t provide robust sales4force automation functionality, but it does suggest where the company,s core competency might still lie "echnical 7unctionality !s your product Web4 based, and if so, howM As described in Chapter 4, Web4based (2* access means all relevant data is accessible via the !nternet 7or those with !nternet access, this is definitely the most fle5ible and secure choice +owever, for remote users who don,t always have Web access, a practical solution might be using applications on their local P(s and subseAuently submitting their changes to a central (2* server 9our reAuirements will tell you which is right for you All the functionality in the world and a really slick user interface don,t matter if you have @> million customers and the vendor,s biggest reference has @ million
+as your product dealt with customer data volumes similar to oursM
+as your product dealt with 9ou might not have pro#ected the number of daily transaction volumes similar to oursM transactions you e5pect your (2* product to handle,
(an you provide a published data model that e5plains all the dataM
!s there a development toolbo5 that Development templates speed up implementation by includes industry or functionally providing programmers with packaged software they oriented templatesM can customi'e rather than writing it from scratch Does the product provide data Although you might not need it now, the fact that the mining or other advanced analyticsM product offers data mining means its functionality goes deep and the vendor has spent time developing advanced functions <or money purchasing them? -ither way, data mining capabilities are a good safety net !mplementation What is your product,s average 3upport implementation timeM "he vendor might lowball this number, so beware Also, understand the difference in time estimates for out4of4the4bo5 configuration versus customi'ation
Does your company help, or do you 3ome (2* vendors are product companies, meaning rely on partners to perform the they partner with systems4integration or consulting partners to perform the workM companies who implement their software for customers /thers have in4house consulting organi'ations that perform implementation duties 0nderstand who will be doing the work !n cases where you rely on consulting partners, who are your partner companiesM *any (2* vendors have long lists of integration partners, allowing you to select the best4Aualified company 0nderstand the vendor,s product certification process and know the skills you,ll be reAuired to invest in And beware the vendor that has only one company e5perienced in customi'ing their product 9ou,ll be at their mercy With a third4party implementation company, knowing where the company,s resources originate will tell you whether you need to factor Btravel and e5pensesC into your development budget and will help you better estimate resource costs
!n cases where you rely on consulting partners, where are those companiesM Where would their staff originateM
After development begins, what do Does the vendor support the development process like you see as your ongoing role with it supports its productM !f the vendor recommends an our pro#ectM alliance partner to assist you in implementation, make sure everyone understands the vendor,s ongoing role 2eferences What percentage of your e5isting customers use the software out4of4 the4bo5 versus reAuiring it to be customi'edM $eyond understanding whether the vendor sees its product as a plug4and4play or highly customi'able, you should understand the relative usage environments between the two <Ask the actual references the same Auestion ?
&erify that the promised functionality actually e5ists -nsure that the product works in the specific technical environment 8auge the product,s usability &erify that the product works with its data
A critical point here is that verifying that the functionality e5ists is one thing, but discovering how the product actually offers the functionality it claims to have is another "wo products might each claim to score campaigns, for e5ample, but one might involve significant end4user input while the other is more automated %ikewise, ensure that the tool can work with your data *any companies ignore this, but it can be a make4or4 break proposition for a (2* program "here might be data problems such as inconsistent formatting that preclude the (2* product from working correctly /r the product might reAuire certain data, such as cleansed address fields or access to customer support history, that your current systems simply can,t furnish Depending on the severity of these data problems, you might want to delay the purchase of any (2* tool until they are resolved -ven if you,re not comparing different (2* tools, actually using evaluation software is a good idea in order to determine whether the per4user cost of the tool is worth the value it provides 7or instance, say you,re a financial services firm that has #ust acAuired an !nternet bank 9ou,re evaluating a call center (2* product that allows seamless integration of customer data from across systems so your (32s have a complete customer profile at the time of the customer,s call "he product costs DF>>> per end4user seat /ne of your evaluation goals should be to verify that the product is able to truly deliver efficiencies that eAual or e5ceed its cost !n other words, if you have I>> (32s across the country, will the resulting productivity gains be worth DI>>,>>> to your companyM "he tool might reAuire a (32 to perform other tasks to get the correct data, mitigating the time saved %ikewise, poor performance might slow down a (32,s ability to resolve a problem /nly by installing the product and testing it can you truly know what to e5pect from it Plan a product evaluation !f you have the time, combine the evaluation with a proof4of4concept that can deliver sample functionality, and demo the functionality to stakeholders for approval (onsider the software evaluation a separate pro#ect with metrics that mirror your e5pectations for the tool after it,s in production Although you won,t be working with all of your data or submitting transactions on the scale you will after the tool is in use, your development team might be able to simulate workloads and e5trapolate performance numbers based on more limited testing At worst, the evaluation will save you time during the actual development pro#ect) at best, it could save you many hours and untold e5pense on a product that doesn,t cut the mustard Checking Re"erences "here,s always one vendor who cites nondisclosure agreements with its customers as a reason not to name names !ndeed, many (2* users consider (2* a strategic weapon and might be unwilling to go on record about what they,re doing !n cases where the vendor uses this as an e5cuse but has cited an anonymous client, ask to have a conversation with someone on the vendor,s account team who can describe the application in general terms Whatever the case, red4flag the (2* vendor who cannot provide you with at least three on4the4record references "he referenced customers should have e5perience developing a pro#ect with the tool, as well as in deploying the tool to end users !deally, they will be in a similar industry or be using the tool in the same capacity6for instance, specifically for e(2*6for which your company is considering it "he actual reference should be a development manager or e5ecutive who has seen the software product deliver what it promised <or not?
"hese Auestions are open4ended and safe "hey invite interpretation, which engages the reference and encourages further e5planation And if the answer to the final Auestion reflects a short4term e5pectation for self4sufficiency, you,re on the right track
Homegrown CR#
Although most (2* development efforts involve some level of software customi'ation, be it to integrate the call center (2* system with marketing,s campaign management software or to change data names, they begin with a core product that provides the foundation from which to customi'e +owever, some enterprising companies have decided to develop (2* software from scratch (ompanies develop their own homegrown (2* systems for the following four main reasons: F = I H "hey reAuire core (2* capabilities that didn,t e5ist at the time they were needed "hey perceive retail packages as being too e5pensive "he combination of core functions they reAuire is too speciali'ed for a single (2* product "hey want assurance of a uniAue solution6one competitors cannot use and vendors cannot reference
Although you might roll your eyes at the conference presenter who claims to have been Bdoing (2* before (2* was even invented,C many companies who have mature data warehouse and business intelligence environments have gradually evolved e5isting customer4focused applications into tailored (2* platforms 3uch capabilities as customer profiling, campaign planning, product affinity analysis, and product recommendation engines might have been deployed to a number of organi'ations across the company at different times $ut because all of the data is centrali'ed on a data warehouse, these companies have formali'ed their processes for integrating and deploying customer data to provide the business with a IK>4degree view of its customers An effective enterprise portal that allows businesspeople a common view of the applications and data means the applications themselves can be distributed and developed at separate times and no one need know !t,s rare, but some companies simply believe that6given (2*,s strategic importance6they have no choice but to build their own (2* environments to maintain their differentiation and their confidentiality A marketing product manager at a large wireless phone company recently e5plained why his company elected to build its own (2* system: "he moment we bring in a vendor to customi'e its software to support our campaigns is the moment that we release our secrets to the competition 3imply put, we sell a commodity product /ur company,s onl& differentiator is in our campaign strategies (all me paranoid, but !,ve seen it too many times ! can,t help but e5pect a (2* vendor to tell our competitors what we,re doing so they can sell to them too, and to use us as a reference "his defeats the entire purpose of automating campaign management, which we,re doing to thwart our competitors6not to clue them in
8sing an A+%
0p until now, this chapter has assumed (2* at your company will be an Bon premiseC solution, that is, built and maintained in4house6as most (2* systems still are "his decision relies on the e5istence of core infrastructure components, including sufficient hardware and software, networking or !nternet capabilities, specific skill sets in both the business and technology areas, and a development process robust enough to deliver a reAuirements4driven program "his is easier said than done for many companies who lack the infrastructure but desperately need to support their customer loyalty or retention programs "hey need (2* fast, faster than the time it will take to hire the necessary staff and install the necessary technology And application services providers <A3Ps? are an increasingly popular alternative BA3PC is often used as a synonym for outsourcing, but A3Ps develop, deliver, and maintain packaged software applications on behalf of their client companies, using the Web as the primary deployment mechanism Although the traditional outsourcing companies speciali'ed in running their customers, BcommodityC systems6billing and human resources being two classically outsourced systems6nowadays A3Ps are bidding on more strategic technology solutions "here are two principal types of A3Ps: F = Web4hosting firms providing customers !nternet access plus a range of services, not to mention a robust technology infrastructure Application providers offering customers access to specific products and product packages
*any pioneering A3Ps began as small specialty shops focusing on specific technologies, most of which are comple5 enough to warrant speciali'ed support +owever, big guys like -D3 and /racle have #umped into the A3P game, offering everything from !nternet service to accounts receivable to campaign management *oreover, the (2* vendors are busy bolstering their own data centers so they can hang out the A3P shingle themselves Why the rush to become an A3PM Web delivery capabilities, for one, are making outsourcing a more realistic choice -nd users accessing (2* from a browser interface need not know whether the customer knowledge is originating downstairs in the data center or three states away at a company they,ve never heard of And (2* is the darling of the A3P community "he 8artner 8roup has forecast that A3Ps will deliver H> percent of all applications by =>>I %ikewise, 7orrester 2esearch has estimated that KH percent of all A3P revenues come from (2* applications With the (2* adoption rate growing e5ponentially, (2* outsourcing is here to stay "he reasons companies elect to outsource their (2* implementation to A3Ps include Ro!u"t technolog& infra"tructure$ (ompanies defining themselves as A3Ps must develop mature technology infrastructures that include robust servers, wide4area networking, operations and database software, application development technologies, and wireless client support for Auick and thorough delivery of multiple software packages /%ee) of im%lementation$ *ost (2* A3Ps usually speciali'e in one or a handful of product packages) thus delivery processes are repeatable across customers, and customi'ation occurs more Auickly E-%erti"e$ A3Ps hire and train their staffs to become e5perts on specific (2* products $ecause these staff members apply a (2* product to multiple user environments, they see the product,s strengths and weaknesses and thus possess an intimacy with the product set that your !" department would struggle to replicate /ervice#level agreement" 0/:A"2$ 3uch contractual agreements establish clear reliability and availability reAuirements to which the A3P must adhere <*any argue A3Ps are more likely to stick with 3%A terms than are internal !" organi'ations ? ritical ma""$ 1ot only is the product e5pertise more solid, but A3Ps also have bench strength, meaning the likelihood of (2* specialist availability is higher "his is especially valuable for
$ut choosing an A3P is hardly a slam4dunk -ach A3P has its own technology specialty, rules of engagement, service levels, and risks Table 8-2 lists some Auestions you should be prepared to ask the A3Ps you,re considering for (2* Table @-7. Buestions to Ask Dour A+% 'he Domain A3P -5pertise 'he +uestion to As ,hy -ou Should As It Do you offer both packaged /ften, companies need a subset of (2* functionality *any (2* suites and standalone A3Ps reAuire their customers to adopt an entire (2* suite (2* modulesM /ther A3Ps, offering standalone customer support, sales force automation, and contract management modules, allow their customers menu4based services +ow many different (2* products do you offerM A3Ps speciali'ing in (2* might offer a myriad of choices, whereas more general Web4hosting firms might offer only one or two "his is a double4edged sword: /n the one hand, the more (2* products a company offers, the more choices you have +owever, the A3P offering a do'en different packages might also be e5pecting too much from its staff in terms of skill sets and availability "he A3P that under4 stands your business is the A3P that can understand customer data hierarchies, for instance, or customer4 focused business processes uniAue to your sector
What are the boundaries of *any A3Ps are simply nuts4and4bolts technology providers !f your servicesM the A3P offers professional services staff, it can provide you with the full system lifecycle, from reAuirements gathering to end4user support +ow solid is your company financiallyM "he financial health of an A3P is a lot more critical than that of your company,s other suppliers 7ind out how the A3P is funded and for how long
"echnical +ow will you integrate your 0nderstand not only how data propagation back and forth from 7unctionality (2* environment with our the legacy systems will happen and how timely it will be <DailyM legacy systemsM WeeklyM? Also understand the lag time between data provisioning and availability (an we e5tract certain data "he A3P might not have the tools or processes that allow the back to our local serverM manipulation or transfer of data outside of the services they provide What type of Auery and Although most A3Ps provide some sort of canned reports, fewer reporting functionality do you provide actual ad hoc reporting capabilities !f this is a
What measures do you have Although the risk of some4 one,s blatantly sharing customer data in place to ensure the with your rivals or other unauthori'ed parties is minimal, there security of our dataM should be written agreements in place to penali'e the A3P for doing so What type of reliability guarantee do you offerM 3ystem up4time is critical in (2* environments, where delays can spell missed opportunities and lost revenues Will the A3P provide =H4by4J supportM When the data is being loaded, will other data be available or will there be outages scheduledM
2eferences
(an we speak to a few "ry to talk to a few customers for whom the A3P has provided companies who have had customi'ation services to assess staff e5pertise as well as timely similar customi'ation needs delivery with the products we,re consideringM
Why build (2* yourself if an A3P can do it all for youM (ompanies foregoing the A3P route cite several valid reasons: (re6uentl& changing re6uirement"* *any companies who have deployed (2* are continually fine4 tuning it and have set up schemes for ongoing end4user input A constant stream of new reAuirements resulting in an updated system might be too much for an A3P who has defined certain standard functionality across its (2* systems and customer base om%le-it& an) timeline""* !ndeed, by the time a company can translate its specific needs to an A3P, it might have been able to customi'e its (2* tool and have it deployed 3o too, many companies insist on being able to test and run marketing campaigns Bon the fly,C whenever they want6a tall order for many resource4bound A3Ps Integration with e-i"ting "&"tem"* "he number and comple5ity of a company,s systems6both legacy systems and emerging technologies6can render data gathering and provision too comple5 and cumbersome for some A3Ps 3o"tage#taking* After a strategic (2* system is in the hands of an A3P, the balance of power shifts 3cary stories proliferate of outsourcers trying to change contract terms while in possession of mission4critical data /ne good BholeC in the contract, and both you and your customers could be at the mercy of an unscrupulous hosting service (ear* With all the efficiencies and speed that accompany the A3P model, some companies are still nervous about relinAuishing their precious customer data into the hands of an outside company, no matter how strident the security policy Plus, many companies who embraced A3Ps early fear training the A3P, who can then repeat the development process for a competitor at a lower cost
"he best A3Ps offer true solutions hosting, combining operational support, change management, and problem resolution for a range of software products gathered into an integrated technology architecture that aligns to your business needs "he problem resolution is so timely and end4user support so seamless your business can,t tell the difference between the (2* provider and your internal !" department "hey give you and your (!/ peace of mind And, most important, they free up your business to concentrate less on technology deployment issues and more on your customers
"he company also uses its data to determine additional services customers might prefer, often deciding to promote their hotel offerings to so4called non4lodgers so customers who routinely %la& at +arrah,s might also "ta& at +arrah,s
The Challenges&
1otwithstanding a fortified !" infrastructure and advanced data analysis capabilities, +arrah,s biggest challenge was convincing individual properties to believe in the newfound customer information enough to
)ood Ad9ice&
3ame4store salesM Product mi5M Profitability4per4visitM +arrah,s managers sound more like a bunch of retailers than gaming industry e5ecs !ndeed, +arrah,s has e5ploited the data in their data warehouse not only to better understand the company,s customers but also to understand overall corporate performance "he company has recently used customer4behavior and financial data from e5isting properties to plan product mi5es and property configurations for newly announced locations, from where to position the buffet to how many and which types of slot machines will be on the casino floor Among other new findings, +arrah,s has used data to determine specific configurations of new properties before they,re even constructed B"he data warehouse has changed the way +arrah,s considers its capital,C says *onica "yson, +arrah,s Director of !" Development for data warehousing BWe,re building a hotel twice as big as we,d planned, based on comparing current properties and their demographics C
A %re-Implementation Checklist
! spend most of my time these days evaluating how pre pared companies are to launch their (2* programs, be they departmental or enterprise4wide, single or multifunction 3ometimes this occurs at the reAuirements definition stage, where there is uncertainty about the perceived need and its implementation viability /ther times it involves evaluating a company,s e5isting infrastructure #ust prior to implementation What ! do most is Aui' key (2* stakeholders about their e5isting environment from both business and technology perspectives *y company calls such evaluations (2* 2eadiness Assessment engagements, but ! like to consider them Bpremortems C After all, what,s more valuable than fi5ing problems before they occurM "he best way to do this is to envision possible outcomes based on current circumstances, using e5periences gleaned from successful (2* deployments !t,s good old risk management, come home to roost Table 9-1 lists a series of considerations to be aware of before moving forward with (2* development *ake sure each of these items has been at least considered at your company, and the more comple5 your intended (2* program, per Table 9-1, the more mandatory it is that you resolve the issue prior to beginning development Table A-1. CR# %re-Implementation Checklist E%aluation +uestion +ave you prepared a (2* business 2egard4 planM E$planation We discussed (2* business planning in Chapter 7 less of whether management reAuires such a document it,s a very good idea to have one that represents (2*,s baseline Considered.
Do you know who your $y the time you,re ready to launch development, the (2* e5ecutive sponsor is and what e5ecutive sponsor should be crystal clear *oreover, her role she e5pectsM in defining and validating reAuirements, managing e5ecutive e5pectations, and helping define success metrics should be well understood by all stakeholders +ave high4level business reAuirements been definedM !n (2* this activity should be separate from the formal development pro#ect for two reasons: business reAuirements will dictate whether the (2* program moves forward, and they reAuire involvement from stakeholders who might not be available during implementation +ow will you know if your (2* program has been a successM Although many companies don,t reAuire success metrics6like
Does each organi'ation agree "he marketing department of an automobile company might on a common definition consider a BcustomerC to be a dealer, but the call center might ofBcustomerCM consider it to be a driver +ave consensus it to be a driver +ave consensus on this and other key definitions before you begin (an you map the desired functionality to data reAuirementsM (ustomer data is comple5 more often than it,s straightforward "his usually means defining data reAuirements along with data reAuirements along with business reAuirements At some point you,ll need to know whether customer data is necessary and from what system it will originate A firm understanding of the level of customer data6account, house hold[1] 6is also critical Purchasing data from an e5ternal source such as Dun L $radstreet, A5ciom, Data Ouick, or -5perian might not initially be a high priority, but it can supplement customer profiles with such indicators as number of family members, estimated income, household4level psychographics, S!P code breakdowns, real estate information, and other attributes that can reveal customer behaviors and preferences What type of workstation con4 figurations does your (2* tool,s development environment reAuireM Additional development tools<e g , *icrosoft,s &isual 3tudio? or hardware <e g , database servers? might be necessary to correctly customi'e the (2* environment "here should be an up4front under4 standing of the impact of (2* on other corporate systems and of how the data will move between systems effectively !n addition, staff members whose systems will be touched by (2* should be notified of the pending integration reAuirements 9es, it,s a loaded Auestion <3ee the end of this chapter ? 1o, it,s not meant to point fingers, but to establish up4front what the tactics will be when Auestions of owner4 ship or disagreements about functional priorities rear their heads An influential e5ecutive sponsor might be able to resolve such issues before they arise 2egardless of whether your (2* program will be Web4based, under4 stand your company,s boundaries for using data about your customers (2* must not only adhere to a corporate privacy policy) it should also be the flagship e5ample of the company,s !ehavior around customer data 3ee Chapter 10 for more about handling privacy
7or customi'ation, does the current workstation development environment support the support the (2* productM +ave you identified the other applications or systems with which the (2* product must integrateM +ave the organi'ational or political barriers to rolling out (2* been identifiedM +ave they been resolvedM
[1]
"he practice of BhouseholdingC organi'es individual consumers into the households in which they live Although the term normally applies to the residential market, business householding groups various organi'ations of a business customer into a common hierarchy "he challenge of householding is getting everyone to agree on the definition of a household
"he most valuable feature of a BpremortemC e5ercise is that it,s a lot easier to give bad news before disaster strikes than to say B! told you soC after the fact6and after the money has been spent (2* assessment findings can alert the business sponsor to potential roadblocks 3uch findings allow (2* team members to
%ead Developer:
Depending on the breadth and comple5ity of your (2* program, the #ob roles listed in Table 9-3 can also participate in (2* development Table A-:. 2ptional CR# 'e9elopment Roles
&ice President of !n large companies, where this position e5ists, the &ice President of 3trategic Planning 3trategic Planning: should be able to share with the (2* team new business areas or product offerings the company e5pects to move toward, acAuisition and partnership strategies, or e5isting products and services the company e5pects to abandon (hief Privacy /fficer: A new position in most companies, the (hief Privacy /fficer should be able to provide details on corporate or regulatory policies regarding the use of customer data
-ach of these #ob roles can play an important part in (2* success, but simply understanding available skill sets can take you a long way in ensuring you can supplement your (2* team with outside help if necessary /f course, such responsibilities as e5ecutive sponsor and the (2* steering committee should be filled by staff members having history with the (2*4related need, pain, or problem, as well as the authority to make decisions "here are roles in (2*, however, particularly in technology implementation areas, where e5ternal e5perts should be leveraged (onsider the following Auestions as you decide whether to beef up your current staff with outside help: 3ow well )o we know the RM ven)or;" )evelo%ment environmentC !t might serve you well to bring in an e5pert from the vendor,s professional services staff or from a partner4integrator to provide knowledge transfer as development gets underway Are there critical one#time#onl& ta"k" that nee) com%letionC 7or work that isn,t likely to be repeated, such as configuring the data, a good consultant can shave days or even weeks off a pro#ect Are we comforta!le that our re6uirement" are well )efine)C 3ometimes an ob#ective third party can find the BholesC in your reAuirements definition "his can help you avoid false starts6which could be a bargain at twice the price an we get "tarte) with our e-i"ting "taffC !t,s often true that by the time you hire and train a full4 time resource, a consultant could have #ump4started a critical task and the entire pro#ect could be that much farther along -veryone would rather hire permanent staff members who have skin in the game, but don,t let principle usurp progress $e willing to focus on the value of time to the business, and invest accordingly "his might mean hiring consultants who can hit the ground running
Another important consideration in (2* staffing is the e5istence of a corporate program management office <P*/? 7irst made popular by the aerospace industry, where comple5 pro#ects were the norm, the P*/ deconstructed a multifaceted system into manageable chunks involving repeatable tasks such as reAuirements definition, software coding, design, testing, validation, and software packaging -ach pro#ect chunk had its own pro#ect manager, goals, budget, and deliverable 0sually stationed in the !" organi'ation, the P*/ is dedicated to running corporate programs such as (2* so pro#ect teams can concentrate on succinct deliverables while the P*/ ties them all together (2* programs are business integration pro#ects whose scope is often corporate4wide <similar to enterprise -2P or supply chain management initiatives? $ecause (2* is driven by business reAuirements and involves the integration of business processes with technology and data, its comple5ity and organi'ational reach is usually greater than the straightforward application "he P*/ formali'es tried4and4true practices that can be applied to (2* implementation "his not only ensures consistency across pro#ects in a program but can also provide consolidated status reporting, often to e5ecutives, affording a level of visibility (2* could never otherwise generate
CR# Implementation
As we discussed at the end of Chapter 7, (2* is usually a corporate program made up of many pro#ects 7or (2* point solutions that deliver finite functionality, one well4run pro#ect might be enough -ach (2* pro#ect should focus on implementing at least one defined reAuirement Whatever the comple5ity, (2* development should be evolutionary and multi4tiered Figure 9-1 describes a departmental (2* program and its associated reAuirements 0igure A-1. CR# program and re-uirements
0nderstanding the comple5ity of your (2* program is critical to planning your (2* pro#ect 7or instance, if (2* is an enterprise initiative, there could be do'ens or even hundreds of discrete reAuirements across the corporation, rendering pro#ect4planning orders vastly more comple5 !f, as in Figure 9-1, the program is departmental, each reAuirement will eventually be deconstructed into a number of different functions, revealing its inherent comple5ity and the development resources it will reAuire
!f, on the other hand, implementation comple5ity is an issue, and the company needs a (2* BAuick win,C the following prioriti'ation might make more sense:
/f course, politics figures into the decision on how to prioriti'e (2* pro#ects After all, if your customer4 support vice president and call4center director are fighting over whether e5ternal data is necessary for really understanding customers, you might want to steer clear of providing (32s with customer profiles until the issue is resolved6no matter how happy it would make the (32s Although formally rating the political landmines of every pro#ect could be overkill6not to mention highly sub#ective6knowing the political baggage that accompanies each potential pro#ect can serve as a tiebreaker When prioriti'ed, a (2* reAuirement6or specific sets of related reAuirements6can be defined as an individual (2* pro#ect as shown in Figure 9-2 0igure A-7. 'elineating CR# proGects
1otice that in Figure 9-2 the Web4related development has been grouped into one pro#ect "his decision was based on practical reasons6specifically, the ongoing challenge of finding available Web4development staff within the company6as well as the estimated development comple5ity Pro#ects F, =, and I are all minimally related and can each leverage e5isting technologies and skill sets within the company Who should scope a (2* pro#ectM !deally, business representatives and development staff should discuss each reAuirement and estimate its value4to4comple5ity ratio6the higher the value and the lower the comple5ity, the better6with the goal of prioriti'ing delivery on an ongoing basis *ost (2* scoping activities focus on delivering initial applications in order to hand over a BAuick winC to the business Applications with a high value4to4comple5ity ratio should rise to the top, and others can be prioriti'ed accordingly "he comple5ity metrics will vary according to the availability of your company,s e5isting technology and staff resources 7or instance, companies that already have robust customer databases won,t rate customer profiling to be as comple5 as those who must start from scratch "o correctly scope a pro#ect, simply rating its functional comple5ity is not enough !deally, you should understand the following: 3pecific technologies that will be involved in implementation 1ecessary skills to implement the pro#ect 1umber of staff members pro#ected to work on the pro#ect 1umber of consultants needed to supplement in4house skills 2ealistic time frame necessary to deliver the first release
3coping a (2* pro#ect prior to launching development mitigates the risks 7or one thing, it,s much easier to develop an accurate pro#ect plan that reflects realistic resource reAuirements, tasks, and time frames Justifying headcount reAuests to management based on the pro#ect,s true scope is also easier 7inally, hiring becomes more straightforward, because the true skills necessary to develop the (2* system are clearer than they would have been if you had simply gone straight to implementation !n fact, failure to thoroughly scope !" pro#ects is one of the principal reasons behind many of their failures
Figure 9-3 illustrates a (2* development roadmap that applies some of this structure 0igure A-:. A CR# implementation roadmap
Within the three main pro#ect phases6planning, construction, and deployment6the (2* roadmap features steps that contain a number of fi5ed and variable tasks: Business %lanning (2* business planning involves many of the steps we discussed in Chapter 7 "he most critical activity at the planning stage is defining (2*,s overall ob#ectives6be they at the department or enterprise level6and delineating the reAuirements of each one At the enterprise level, (2* business planning can involve the documentation of a corporate (2* strategy and the definition of the corresponding programs within it At the department level, it can simply mean establishing the boundaries of a new (2* application At minimum, the business4planning phase should include the documentation of high4level (2* business goals in the form of a strategy document or business plan "his document will be leveraged at (2*,s inception to gain e5ecutive consensus and sponsorship !t will be useful as a focal point for reAuirements4 driven development and6after the (2* pro#ect has deployed an application6as a way to measure its results As Chapter 7 illustrated in the hotel reservation system discussion, part of business planning should identify the critical customer4focused business processes (2* will impact Where they are straightforward, you might decide to redesign these processes as part of the planning activity *ore often than not, companies planning their (2* pro#ects reali'e that rather than simply auto mating e5isting business processes, they are defining those processes for the first time Depending on funding and sponsorship reAuirements, (2* business planning might optionally include 2/! estimation or cost4savings pro#ections Architecture and 'esign "he need to plan (2* architecture and to design an implementation strategy is what makes business sponsors and pro#ect leaders shudder and go straight to technology selection hoping for a miracle "he architecture and design step is painful, but it,s worth it "his step identifies the business processes the (2* product will support !t involves listing the specific functions that will need to be implemented6and how6ultimately giving you a good idea of (2*,s impact on the organi'ation and various technologies !nventorying the range of corporate areas (2* will affect, as well as those that will affect (2*, is a critical activity At the end of this step you should be able to answer the following two Auestions: F = What technologies and processes do we have in place that will be impacted by (2*M What do we need that we don,t have today in order for (2* to workM
2elative to e5isting technologies, try to pro#ect (2*,s impact on your current systems 9our !" organi'ation should be willing to do this6and in return it won,t be blindsided by (2* after it,s been developed !mpact analysis can mean listing current systems6for instance, you might need to know a bit about your company,s e5isting call center operational system before you can understand how candidate (2* technologies will link to it !ndeed, a range of e5isting technologies, from -2P systems to current marketing automation technologies to handheld computers, are likely to be touched by (2*
"his letter was included in a product delivery and represents a veritable smorgasbord of (2* don,ts "he first one is that the company,s online ordering system is obviously not linked to its inventory system <"he fact that the company happens to be a high4profile dot4com with an edgy Web site and slick e4mail marketing campaigns is not evident in its post4sales customer support ? "he customer should have been notified of the out4of4stock items at the time of the order, not upon delivery of the remaining items "he company might believe that, had the customer known that not all of the items she ordered would be in stock, she would not have placed the order at all Perhaps some of the out4of4stock items the customer wanted were in some way related to the items that showed up on her doorstep /r perhaps the company
-A! is important to (2* because, no matter how successful a new marketing campaign or how polite the <albeit mass? marketing message, if internal systems cannot share data, vital business knowledge could be lost and customer service undermined !f the company truly had -A!, its inventory system could alert its customer support system when the desired items came back in stock, allowing a (32 to notify the customer and make a sale !t is for this reason that many companies undertake -A! as a preparatory step toward (2* 7or new (2* functionality, you,ll also need to understand what data to consider 7or each business reAuirement, one or more data reAuirements will result 7or instance, if survey data is to be incorporated into customer profiles, which specific data elements should be collectedM Will you need to collect e5ternal data such as third4party householding information or competitive intelligence dataM /f the data collected, what should be displayed to (32sM "o marketing staffM And what systems will deliver that dataM A significant part of defining data reAuirements involves addressing the actual meaning of certain data definitions !s there consensus across the business that the term BrevenueC means booked revenue, or might it imply billed revenueM Does a Bnew customerC have the same attributes in the sales organi'ation as in customer supportM "o many in !", documenting data definitions smacks of cumbersome metadata management and documentation pro#ects +owever, it,s more about simply gathering consensus and enforcing consistent business terminology, whatever form that takes !f information is indeed a corporate asset, as we discussed in Chapter 7, consistent and sustainable data definitions are essential When you,ve completed an impact analysis, you can begin prioriti'ing pro#ects according to business reAuirements and staffing your development pro#ects, as we discussed earlier in this chapter Technolog! +election As Chapter 8 e5plained, (2* technology selection can be as simple as choosing an off4the4shelf product or as comple5 as a comprehensive evaluation of various (2* systems integrators or A3Ps !f you,ve bitten the bullet during architecture and implementation design, understanding (2*,s impact on e5isting systems and its reAuirements for new functionality, you should be in good shape to align any candidate (2* product to your e5isting !" environment 'e9elopment Development involves the construction and customi'ation of the (2* product, using specific product features $ut (2* development is more than programmers assuming center stage and writing code) it involves the integration of business processes with the chosen (2* product $y this time, you will have already identified the key (2* business processes Process integration means that (2* technology you,ve #ust selected integrates into these business processes <"he converse6 merging business processes into the (2* product,s features6forces the product to in effect define or change those processes, thereby diluting them until they are no longer optimi'ed ? Process integration involves ensuring that identified business processes are tested with users to ensure not only that the business processes work, but also that technology features can be leveraged in order to refine them !n other words, technological capabilities should improve, not compromise, customer4focused business processes 7or instance, a campaign management product allows segment managers the opportunity to refine a mailing list before the campaign is launched6something they,ve never been able to do6thus refining the e5isting process "he same product might also allow a campaign director to monitor a campaign,s success rate as it,s being e5ecuted !f the first thousand prospects have been unresponsive, the manager can cancel the campaign rather than allowing it to proceed, adding another valuable option to the campaign e5ecution process 2efining business processes during development means iterative proto typing: from time to time programmers demonstrate interim functionality to business users "hus business users can monitor product
(ustomer support cross4 (ustomer support selling increase of K cross4selling increase percent of F> percent
-limination of sales staff responsibility to produce weekly and monthly reports, generating a productivity increase of @WF> percent -limination of need for data correlation by implementing centrali'ed customer database
FF percent increase in F= percent increase in sales productivity and sales productivity reduction of one full4time administrative position 1one6database pending -limination of data correlation, resulting in redeployment of two full4time data analysts
*easurement also includes the incorporation of user feedback to improve (2* usability and business effectiveness As the (2* implementation roadmap shown in Figure 9-3 illustrates, (2* measurement loops back around to further (2* business planning, allowing the company and its lines of business to continually refine (2* reAuirements and identify new (2* opportunities at the same time !f you incorporate measurement and feedback into the planning cycle, (2* will deliver new and better functionality, resulting in small victories that add up to improved customer value
'on't&
!t,s unlikely that any e5ecutive will be ready to sign on the bottom line after simply hearing about (2* for the first time As with everything else, selling (2* internally is a process, and the first step,s always the hardest $e prepared to educate people and, at worst, to begin chipping away at some closely held paradigms !n many companies, advocating a newfound customer focus is tantamount to heresy 3ome people will need time to evolve from roadblocks into hardcore (2* constituents
CR# Roadblocks
3peaking of roadblocks, there are lots of them on the #ourney to (2* success 3ome obstruct the beginning of the pro#ect during the business planning or reAuirements gathering that allows a (2* team time to set things right *ost (2* roadblocks aren,t single events, however, but processes that gradually erode a (2* effort until the ultimate system becomes a mere shadow of its intended vision
The 0our %s
A good (2* pro#ect can go bad for many reasons $ut the principal factors6! call them the 7our Ps6are Process, Perception, Privacy, and Politics %rocess /ne of the failures ! encounter most often in my (2* fieldwork is with companies that are slow or unwilling to modify their business processes to support better customer relationships 3ome are simply unwilling to acknowledge that their business processes warrant improvement As we discussed in Chapter 7, the weaknesses of e5isting business processes, such as accepting a hotel reservation from a freAuent guest, should be clearly understood and refined prior to implementing (2* technology
Whether or not they differentiate the overblown scare tactics from the legitimate threats, the vast ma#ority of Web surfers favor punishing companies6and even top e5ecutives6for privacy violations [2] While the debate rages, what can you do to ensure that pending privacy regulations don,t foil your (2* programM
[2]
According to a report in "he3tandard com <August =I, =>>>? citing a Pew !nternet and American %ife Pro#ect study, GH percent of Web surfers surveyed are proponents of punishing companies and their top e5ecutives for privacy violations
0nderstand the 6ui) %ro 6uo$ consumers are more likely to share their personal information with your company if they receive something valuable in return <7or instance, travelers reserving their flights on -5pedia com are e4mailed weather reports for their destinations a few days before they fly ? !ncorporate this into your (2* planning to ensure that customers are sufficiently motivated to continue interacting with you at every touchpoint Post your privacy policy conspicuously on your Web site, as well as in print ads and other customer communications 0nderstand permission marketing and the trade4offs between asking customers to opt in versus opt out Decide which approach is best for each discrete marketing channel and, if practical, for every customer (onsider giving your customers full access to all the information you have about them Although admittedly risky, this tack makes the customer a participant in managing his information, potentially enhancing his perception of your company as a partner 7or instance, Allstate customers can log on to a secure Web site and actually change their own profiles, benefiting both parties
%ike (2*, privacy is big business /nline marketing companies such as *atch%ogic and Double(lick <whose stock price tumbled when it revealed plans to identify individual consumers? base their operations on tracking anonymous clickstreams that have been cleansed of their consumer4specific data %arge consulting firms are offering privacy audits6often at hundreds of thousands of dollars a pop And a rash of legal firms speciali'ing in online privacy law have sprung up virtually overnight (onsumer watchdog groups such as the -lectronic Privacy !nformation (enter <www epic org? and Junkbusters <www #unkbusters com? aren,t letting up) they continue to publish consumer alerts and push for stricter regulatory measures And www privacy org, whose slogan reads BPrivacy is a 2ight, 1ot a Preference,C regularly features corporate privacy breaches in its headlines !n the meantime, businesses with a stake in consumer data are rapidly forming alliances to ease consumer fears while warding off the threat of governmental regulation "he 2esponsible -lectronic (ommunications Alliance <www responsibleemail org? recently created a set of self4regulatory guidelines for e4mailWbased marketing campaigns, and "he Personali'ation (onsortium <www personali'ation org? is a business advocacy group dedicated to illuminating the path toward responsible one4to4one marketing practices And not a second too soon6visionary companies who once proudly heralded their one4to4one marketing capabilities and detailed customer databases are increasingly clamming up for fear of another privacy backlash $ut marketing strategy in the back room instead of in the boardroom will do little to comfort wary consumers (ircumspect and deliberate planning of customer interactions and data4gathering will do a whole lot more %olitics *y company has a long4term relationship with a ma#or 0 3 financial institution, a client whom we,ve accompanied on the (2* #ourney from planning through e5ecution and measurement and back to planning ! know the struggles the company has had and how it has effectively worked through solutions to deliver (2* to a hungry group of private bankers whose compensation is based on a customer retention and satisfaction scorecard 7rom time to time ! attend conferences featuring presentations from representatives of this firm 3ome of the presenters ! know) others are strangers /ften they discuss (2* pro#ects !,ve never heard of, deployed to lines of business who weren,t involved in the sanctioned, corporate4wide (2* program our team helped deliver After sitting through diverse and often unrelated (2* presentations from various employees of this client, ! know one thing for sure: lurking in the farthest reaches of the headAuarters campus are BskunkworksC (2* pro#ects6pro#ects that are Auietly being developed, using leftover budget money or developers unaccounted for on other pro#ects 3kunkworks (2* pro#ects are usually a sign of urgent business needs combined with a high degree of cynicism about delivering (2* via the appropriate channels <e g , the !" department? "hey can also mean that certain organi'ations or individuals in need of (2* functionality might be unaware of (2* activities occurring elsewhere in the company *ore insidiously, skunkworks pro#ects can sometimes indicate nasty politics 7rom time to time a rogue manager, unhappy about the way the legitimate (2* program is playing out or disillusioned with the pro#ect team,s leadership, might decide to go it alone and build her own (2* environment 3he meets with (2* vendors, several of whom lost their bids for the sanctioned (2* activity, and hoards an e5clusive set of analyst research reports Although e5cuses range from B! need to use the budget money or !,ll lose it,C to B!t,s not really (2*) it #ust uses the same data,C the potential to dilute the value of the company4endorsed (2* program is real Another freAuent political scenario involves the development of a data warehouse or other (2*4related technology solution and labeling it (2* without defining a clear (2* strategy, planned process
'emanding Customers !t,s not your fault 9ou were trying to capture market share 3o now your repeat customers have come to e5pect D=> off on every D@> purchase What do you do nowM !t might be too late, at least for that seminal group that first found your Web site and continues to respond to e4mail promotions and cherry4pick at every turn $ut it,s not too late for the ne5t batch, who might buy from you again because your site is easy to navigate and your customer support friendly and timely !f you,re considering offering great deals to rope in customers, reali'e that attracting them doesn,t mean keeping them 0nderstand how to identify your negative4value customers, and avoid enticing them with money4losing deals 0nderstand who they are when they do show up, and offer them opportunities to become profitable $ut also know your high4value customers and differentiate how you treat them, whether they,re entering your store or your storefront Customer +er9ice That's Reall! Bad /ne ornery call center rep who drones on about company policy and won,t try any workarounds on behalf of a flummo5ed customer can effectively eradicate, in G> seconds, the toil and cost of an entire (2* initiative Chapter 3 discussed the deservedly high profile of the customer contact center and its ability to make or break customer loyalty Poor customer service comes in many guises !t,s the employee at the home improvement warehouse whose answer is B1oC before you even finish your Auestion, the salesperson who is unable to take your order because of system problems, and the fourth consecutive B3erver is downC message displayed when you try to access an e4tailer,s Web site $ut poor customer service can also mean aggressive telesales staff, chronically late suppliers, or a sales rep too busy to take your call now that you,ve placed your order !t can also mean the well4meaning employee who has taken the customer relationship management strategy to heart, as in the cartoon
Whatever the problem, implement a series of new customer service guidelines for every customer4facing department in your company *ake these guidelines part of staff members, performance evaluations and compensation packages !n short, foster ownership of customer satisfaction !t,ll be money well spent
7rom B$usiness Defined: (onnecting (ontent, Applications, and (ustomers,C E.8 a% @emini1 =>>>
"he privacy debate has been partly responsible for a growing backlash against all forms of database marketing, however innocuous Privacy groups are admonishing consumers to enter wrong data each time they fill in an online profile or enter a direct4mail sweepstakes, thereby foiling companies, efforts to use their personal information <if for nothing more than to understand who they are? And because multiple family members usually share one household P( and your daughter has #ust bought her third *arilyn *anson (D from an online music retailer, #ust how accurate are the preferences being captured by those ubiAuitous cookiesM !f a (2* backlash happens, it probably won,t begin in the home offices of wary Web users but in the e5ecutive suites of ma#or corporations -5pectations for what (2* can deliver are ever more lofty as software vendors make grandiose claims and (2* press releases make news !n June of FGGG, 7orrester 2esearch published the provocatively titled research paper B"he Demise of (2*,C doubtless targeted at vulnerable e5ecutives who had begun Auestioning whether (2* was really worth it !ncreasing e5pectations for (2* invite the possibility that its cost could outweigh its ultimate value to an organi'ation $y the time you,ve reali'ed (2*,s promise, determined which organi'ations can benefit from it, convinced them to work together, determined the data reAuirements, loaded the database, integrated disparate applications, and customi'ed the software, (2* could be costing you millions And this is before it,s even been deployed A noted outdoor gear retailer with both an established brick4and4mortar presence and an acclaimed Web site recently told an !" #ournal that it has only two customer segments6one for Bprofessional shoppersC <people who like to shop for multiple items and coordinate outfits? and another for shoppers who know e5actly what they want to buy "he company maintained that understanding these two segments gave it the necessary information to craft effective marketing messages "his company is not an upstart but an established retail presence with a loyal customer base !t,s regularly lauded for offering the latest outdoor apparel at a fair price for a broad range of consumers !f ! were a (2* salesperson, ! wouldn,t want them in my territory !f your company sells a commodity4based product, the ability to gain additional market share through (2* might not be worth the investment "here might be better ways to acAuire customers, such as dropping prices, hiring additional salespeople, or adopting additional channels to gain market share 7or these types
Appendi Bibliograph!
%urther Reading 0urther Reading
7ollowing is a list of resources for you to supplement your knowledge of (2* and its various components !,ve broken down the material into three categories: books, maga'ines and #ournals, and Web sites Books "hese books cover a range of (2* and (2*4related topics Adriaans, Pieter, Dolf Santinge Data Mining 2eading, *A: Addison4Wesley, FGGK $arlow, Janelle, Dianna *aul Emotional +alue$ reating /trong Bon)" with 8our u"tomer" 3an 7rancisco: $errett4.oehler, =>>> $erry, *ichael J A , 8ordon %inhoff Data Mining Techni6ue"$ (or Marketing1 /ale"1 an) u"tomer /u%%ort 1ew 9ork: John Wiley L 3ons, FGGJ $rown, 3tanley A u"tomer Relation"hi% Management$ A /trategic Im%erative in the Worl) of E#Bu"ine"" 1ew 9ork: John Wiley L 3ons, =>>> Dych, Jill e#Data$ Turning Data into Information with Data Warehou"ing $oston: Addison4Wesley, =>>> 8ladwell, *alcolm The Ti%%ing 'oint$ 3ow :ittle Thing" an Make a Big Difference $oston: %ittle, $rown, =>>> 8odin, 3eth 'ermi""ion Marketing$ Turning /tranger" into (rien)" an) (rien)" into u"tomer" 1ew 9ork: 3imon L 3chuster, FGGG 8ordon, !an Relation"hi% Marketing$ 5ew /trategie"1 Techni6ue"1 an) Technologie" to Win the u"tomer" 8ou Want an) =ee% Them (orever 1ew 9ork: John Wiley L 3ons, FGGE +ughes, Arthur /trategic Data!a"e Marketing* /econ) e)ition 1ew 9ork: *c8raw4+ill, =>>> !mhoff, (laudia, %isa %oftis, Jonathan 8 8eiger Buil)ing the u"tomer# entric Enter%ri"e$ Data Warehou"ing Techni6ue" for /u%%orting u"tomer Relation"hi% Management 1ew 9ork: John Wiley L 3ons, =>>F !nmon, William + Buil)ing the Data Warehou"e 1ew 9ork: John Wiley L 3ons, FGGK !nmon, William + , 2ichard D +ackathorn 9"ing the Data Warehou"e 1ew 9ork: John Wiley L 3ons, FGGH .alakota, 2avi, *arcia 2obinson E#Bu"ine"" 2*<$ Roa)ma% for /ucce"" $oston: Addison4Wesley, =>>F .imball, 2alph, %aura 2eeves, *argy 2oss, Warren "hornthwaite The Data Warehou"e :ifec&cle Toolkit$ E-%ert Metho)" for De"igning1 Develo%ing1 an) De%lo&ing Data Warehou"e" 1ew 9ork: John Wiley L 3ons, FGGE %evine, 2ick, (hristopher %ocke, Doc 3earls, David Weinberger The luetrain Manife"to$ The En) of Bu"ine"" A" 9"ual (ambridge, *A: Perseus Publishing, FGGG %inthicum, David 3 B2B A%%lication Inte gration$ e#Bu"ine""IEna!le 8our Enter %ri"e $oston: Addison4 Wesley, =>>F *artin, (huck 5et (uture$ The H &!er tren)" that Will Drive 8our Bu"ine""1 reate 5ew Wealth1 an) Define 8our (uture 1ew 9ork: *c8raw4+ill, FGGG 1ewell, 7rederick :o&alt&*com$ u"tomer Relation"hi% Management in the 5ew Era of Internet Marketing 1ew 9ork: *c8raw4+ill, =>>> Peppers, Don, *artha 2ogers The One to One (uture$ Buil)ing Relation "hi%" One u"tomer at a Time 1ew 9ork: (urrency Doubleday, FGGI [[[[[[[[[[ Enter%ri"e One to One$ Tool" for om%eting in the Interactive Age 1ew 9ork: (urrency Doubleday, FGGJ [[[[[[[[[[ The One to One Manager$ An E-ecutive;" @ui)e to u"tomer Relation"hi% Management* 1ew 9ork: 2andom +ouse, FGGG Pine, Joseph, James 8ilmore The E-%erience Econom& $oston: +arvard $usiness 3chool Press, FGGG 2eichheld, 7rederick 7 The :o&alt& Effect$ The 3i))en (orce Behin) @rowth1 'rofit"1 an) :a"ting +alue $oston: +arvard $usiness 3chool Press, FGGK 3eybold, Patricia $ u"tomer"*com$ 3ow to reate a 'rofita!le Bu"ine"" /trateg& for the Internet an) Be&on) 1ew 9ork: 2andom +ouse, FGGE 3eybold, Patricia $ The u"tomer Revolution$ 3ow to Thrive When u"tomer" Are in ontrol 1ew 9ork: (rown $usiness, =>>F
5eb +ites
*ore and more (2*4related Web sites are popping up every day ! focus here on those that are heavy on content and light on advertising and vendor sponsorships +owever, with any Web site that markets itself as a repository of information, especially with those that offer product reviews, you would always do well to note the sponsors who are footing the bill as you review products, read vendor case studies, and choose (2* conferences and workshops +aving said that, don,t dismiss the Web sites of (2* product vendors *any feature white papers and (2* news releases along with information on their specific (2* solutions www !r%!$%%u"it) !$%: /ffers a range of (2* resources Provides an especially thorough compendium of white papers on all topics "he Bconsultants cornerC that allows visitors to ask Auestions of (2* e5perts and a chat room are good places to monitor current (2* vendor and product bu''
www !r%2001$"li"e !$%: A virtual (2* conference, this clever site lets visitors attend seminars and wander through a virtual trade show $ooths feature vendor presentations and provide links to (2* vendor Web sites &isitors can even collect brochures in their virtual shopping bags C)beratla# i"ter"et !$%: 7or the latest news on e4commerce companies "he place to go when you want to see who,s bought whom www *e#ti"ati$"!r% !$%: (2* product reviews, analyst reports, and surveys (overs everything from A3P reviews to wireless technology trends to vendor coverage www *##tar !$%: 1ews and editorials about business intelligence, data warehousing, and data mining www *wi"&$!e"ter $rg: Provides information on a range of data warehousing and business intelligence topics /b#ectively lists data warehouse vendors and consultants, breaking them down according to their focus Also has links to various publications featuring data warehouse news and editorials www *w-i"#titute !$%: "he Data Ware housing !nstitute,s Web site, featuring articles, white papers, case studies, and resource guides, as well as updates about its conferences "DW! is the only conference that does not cull speakers from its list of e5hibitors, thus bestowing real4world e5periences in data warehousing and business intelligence www e!r%gui*e !$%: A repository for (2* news and vendor press releases, updated several times a day www epi! $rg: 8rassroots lobbying effort for 0 3 4based privacy legisla tion, this site has even proffered editorial viewpoints on various politicians, privacy track records Accompanying subscriber newsletter broadcasts regular briefings on privacy developments www itt$$lb$/ !$%: A knowledge center for all things !" /ffers a Btoolbo5C for both a (2* and $usiness !ntelligence <among other areas?, allowing visitors to review product features and read case studies about their applications "he site also offers subscriptions to free newsletters for managers and practitioners "he site,s (2* area does a particularly thorough #ob of deconstructing (2* into its various components ipw i"ter"et !$%: 2un by !nternet Product Watch, a comprehensive clearinghouse for e4business vendors and products www 0u",bu#ter# !$%: Web epicenter of privacy legislation and protection !ncludes freAuent updates on legisla tive measures in the 0nited 3tates and follows privacy4related lawsuits Provides downloadable software and e4mail addresses for blocking the use of cookies and other methods for data gathering www $pe"it/ !$%: B/pen !" -5changeC offers insight into a range of e4business areas, from knowledge management to operating systems to data management "he site also offers a weekly newsletter to subscribers www per#$"ali+ati$" $rg: "he name belies this site,s purpose, which is to discourage privacy legislation in favor of corporate responsibility Also offers news and press releases on the personali'ation and relationship marketing industries www pri'a!) $rg: /ffers articles, news releases, and links to other Web sites with the aim of pushing privacy legislation in the 0nited 3tates +as been known to publici'e companies selling or sharing customer data e5ternally www #ear!h!r% !$%: Presents a variety of (2* resources under one virtual roof !ncludes an introduction to (2* basics as well as a host of features for visitors looking for product comparisons, book reviews, current articles, and white papers <both original and gleaned from other publications and Web sites? !ncludes a (2* #ob board
)lossar!
Abandoned shopping cart&
3et of products a customer intended to buy, but did not ultimately purchase Although possible in a brick4and4mortar store, the term more accurately applies to a Web site where a visitor collected one or more items in his virtual basket but never made it as far as checkout
A""init! anal!sis&
Detecting sets of products or services purchased together -5ample: "ortilla chips and salsa
Anal!tical CR#&
0se of data originating through front4office or operational (2* to enhance customer relationships (ombined with other organi'ational or e5ternal data to evaluate key business measures such as customer satisfaction, customer profitability, or customer loyalty to support business decisions
Attrition&
(ustomer leaving to go do business with a competitor 3ee also [Churn:]
Automated work"low&
-nabling work processes to BflowC through a company without human intervention Workflow systems usually involve moving data through a process, such as order and fulfillment, that reaches across various systems and departments
B7B&
(ommon abbreviation for Bbusiness4to4business C
B7C&
(ommon abbreviation for Bbusiness4to4consumer C
Back-o""ice CR#&
"he area of (2* that involves analysis to optimi'e customer4facing business processes and revenues 3ee also [Anal&ti'al CRM:]
Brick-and-mortar business&
A physical storefront or branch $rick4and4mortar can be a company,s core business or simply one of several sales channels
Business intelligence&
1ormally describes the result of in4depth analysis of detailed business data !ncludes database and application technologies, as well as analysis practices 3ometimes used synonymously with Bdecision support,C though business intelligence is technically much broader, potentially encompassing knowledge management, enterprise resource planning, and data mining, among other practices
Business sponsor&
*anager or e5ecutive who acts as visionary for the (2* program and can articulate how (2* can drive business improvements "his person establishes the Bneed, pain, or problemC (2* will solve, serves as a tiebreaker for issues during the pro#ect, and might actually fund some or all of (2* development
Call center&
/rgani'ation in charge of direct customer support interactions "he term Bcall centerC refers to the classic telephone support infrastructure and is being replaced by Bcontact centerC or Bcustomer care center,C both of which imply more technological sophistication and multichannel support
Call routing&
Directing customer calls to a specific agent based on a specific parameter such as that agent,s e5pertise, geographical location, subsidiary affiliation, or other characteristic
Call scripting&
3ee [('ripting:]
Campaign management&
Analy'ing data for purposes of launching a marketing campaign and then monitoring that campaign and tracking its results to determine the campaign,s value (an also refer to the technology that automates the campaign manage ment function
C-commerce&
3ee [Collaborati!e 'ommer'e:]
Channel&
*eans and media by which a customer prefers to communicate with the company <the BinboundC channel? or to receive communications <the BoutboundC channel?
Channel optimi3ation&
Determining the best channels by which to communicate with and sell to customers, especially when they have not made these preferences clear, and making these channels available to the right customers 7or e5ample, a company might choose to use resellers to sell a product in order to minimi'e its costs
Churn&
(ustomers leaving your business to go to a competitor !mplies the customer might or might not return B(hurn reductionC is another way of saying customer retention and is a ma#or goal of (2* (hurn is most often used in con#unction with highly competitive commodity product businesses such as communications companies, utilities, and airlines
Clickstream&
3eries of page visits and associated clicks e5ecuted by a Web site visitor while navigating through the site Analysis of clickstream data can help a company understand which products, Web site content, or screens were of most interest to a given customer
Collaborati9e commerce&
.nown as Bc4commerce,C reflects the ability of various partners within a supply chain to share important data about products, inventory levels, and orders
Collaborati9e CR#&
3pecific functionality that enables a two4way dialog between a company and its customers, through a variety of channels, to facilitate and improve the Auality of customer interactions
Con"iguration support&
0sually a com ponent of sales force automation Automates the estimation process for si'ing and pricing a product, using prospect or client data as input (ompanies selling comple5 custom products, such as computer eAuipment, are the main users of configuration support tools
Con"igurator&
(ommon name for tool that performs configuration support Also the brand name for 3iebel,s configuration support tool
Contact center&
A more sophisticated version of the classic call center, which was staffed by telephone operators (ontact center suggests a greater degree of technological sophistication, including multimodal customer support, outbound telemarketing, and customer self4service
Contact management&
Area of sales force automation that allows salespeople to record key customer information such as names and addresses, as well as organi'ation charts and account activities Prevents salespeople from having to remember who,s who at each of their accounts
Control group&
8roup of customers, usually randomly selected, whose responses to a campaign are compared to those of a specially selected customer group that is more systematically chosen -valuating customer responses to a specially designed communication with the responses of a control group can indicate the effectiveness of a target marketing campaign
Cookie&
0nit of te5t placed on someone,s computer when he accesses a Web site, intended to serve as a permanent way for the site to recogni'e that customer when he returns Allows the site to track various pieces of information the customer submits to the site across multiple visits
Cross-"unctional&
Description for a technology that serves more than one business function <e g , financial analysis and sales analysis? or organi'ation <e g , +uman 2esources and Accounts 2eceivable? (2* point solutions, by their definition, are not cross4 functional, whereas (2* suites might be
Cross-selling&
3elling a customer a product or service based on her past behaviors or purchase history $est done when a company understands the relationship between two products and identifies which product might BpullC another
Customer segmentation&
3ee [(egmentation:]
Customi3ation&
(ustomer,s ability to tailor Web site content to her specific needs, interests, and usage preferences 3ee also [Personali)ation:]
C!beragent&
3oftware program that can guide a user,s decisions and recommend potential action (yberagents can help customers navigate a Web site, answer freAuently asked Auestions, and suggest ne5t steps, and their capabilities are growing all the time 3ome cyberagents are animated to appear human
'ata mining&
Advanced analysis used to determine certain patterns within data *ost often associated with predictive analysis
'ata mart&
0sually refers to a physical platform on which summari'ed data is stored for decision support Data marts are commonly used for specific analysis purposes by a single organi'ation or user group
'ata warehouse&
(ollection of integrated data used for decision4making 1ormally the system of record for detailed customer data from heterogeneous systems across the company to provide a consistent view of the business
'ecision support&
Data analysis with the purpose of fueling accurate and effective business decisions .nown by the abbreviation BD33,C for decision support systems, decision support usually involves accessing data on a data warehouse
'irect marketing&
(lassic marketing practice of communicating directly to consumers, normally via the postal service +as evolved to encompass a range of media, from e4mail to banner ads to wireless messaging services
e-marketplace&
/nline e5change that enables buyers and sellers in a supply chain to come together, providing better information <better than they would have with classic human4 intensive supply chain processes? as well as automation of key business processes
4nterprise CR#&
(ross4functional (2* system used across various organi'ations and departments A salesperson reviewing his customer,s most recent open trouble tickets before making a sales call is an e5ample of enterprise (2*
4nterprise portal&
0ser interface, usually Web4based, that provides a virtual window into different data sources and sub#ect areas Provides a common look and feel across the company for various business data, but can nevertheless be customi'ed to provide user4specific information at various levels and from various sources
e-tailer&
B!nternet retailer C 0sually refers to a company for whom the !nternet is the e5clusive sales channel
49ent-based marketing&
Detecting a key event that triggers a tailored marketing communication or business action designed to increase customer loyalty or profitability -5ample: 2esponding to a customer,s inordinately large bank deposit with an offer for a high4interest certificate of deposit
4 change&
(ompany partnership that leverages spending power and critical mass to negotiate more favorable deals with suppliers 0sually specific to a single industry and often involving companies that compete as well as cooperate with each other
4 ternal data&
Data acAuired outside of a company,s internal !" organi'ation 0sually entails consumer data that has been cleansed, formatted, and updated with current information, but might also include market research, demographic statistics, and business and industry information
4 tranet&
3ecure !nternet site available only to a company,s internal staff and approved third4party partners 7lourishing in $=$ environments where suppliers can have ready access to updated information from their business customers, and vice versa
0ront-o""ice CR#&
(ustomer4facing (2* capability 0sually pertains to sales force automation systems and other systems that involve direct customer inter actions that can be recorded for back4office analysis
Householding&
(onsolidating customer data to organi'e individuals into the households in which they live 8rouping customers within a household allows a company to be more prudent with its communications and at the same time to more accurately profile individuals in relation to one another
Incremental de9elopment&
Deploying periodic releases of software such that end4users receive functionality Bin chunksC and not all at once
In"erential personali3ation&
0sing analysis and e5trapolation of customer behaviors and preferences, including performing collaborative filtering and other types of data mining, to tailor Web content and dictate the optimal marketing message
Iterati9e de9elopment&
Desirable (2* approach of going through small, repeatable development steps to speed up software implementation and deliver small amounts of functionality more Auickly 2educes risk and allows for ad#ustments, enabling technical staff to refine the development plan as they go
,ead management&
3ales force automation capability for tracking and monitoring sales prospects and a company,s interactions with them, as well as enforcing sales tactics and automating key tasks Also enables telemarketing staff to pass leads along to the appropriate sales channel
,i"e-stage marketing&
"argeting consumers based on where they are on their life continuum 7or instance, a bank might e4 mail a promotion for a new credit card to a recent college graduate, but might market a home eAuity line of credit to a recent home buyer or financial planning to a recent retiree
,ist generation&
,oad balancing&
3ee [Automati' 'all distribution:]
#arket-basket anal!sis&
Analysis of items purchased together during a shopping trip 1eed not be retail specific but can also be applied to a bank,s or telephone company,s products (lassic e5ample: Peanut butter being purchased with #elly
#ass marketing&
"raditional practice of marketing a product to an undifferentiated group of consumers Also known as Bspray and prayC or Bbatch and blast C
mCR#&
*obile (2* (ommunicating key information to customers or internal customer support staff via wireless technologies
#etadata&
BData about data C 0sually refers to agreed4on definitions and business rules stored in a centrali'ed repository so business users6even those across departments and systems6use common terminology for key business terms (an include information about data,s currency, ownership, source system, derivation <e g , profit X revenues minus costs?, or usage rules Prevents data misinterpretation and poor decision4making due to a sketchy understanding of the true meaning and use of corporate data
#ultichannel&
3upport of more than one sales or service channel) for instance a retailer,s Web site and its catalog
#ultimodal&
1umerous ways a customer can interact with a company, and vice versa A call center that allows inbound customer interactions via fa5, voice, telephone keypad, or handheld device is providing multimodal access
2perational CR#&
2pportunit! management&
3ee [+ead management:]
%ermission marketing&
A customer,s implicit or e5plicit agreement to be communicated to or to communicate with a company 0sually implies customer perception of value in the relationship, suggesting a 6ui) %ro 6uo between the customer and the vendor
%ersonali3ation&
(ompany,s ability to recogni'e a customer or prospect as an individual and differentiate its inter actions with her Although personali'ation usually means individuali'ed content delivered on a Web site, it can also involve target marketing, tailored e4mail campaigns, or customi'ed banner ads
%ipeline #anagement&
3ee [+ead management:]
%ri9ate portal&
3ee [,-tranet:]
%oint solution&
Piece of software used for a specific business purpose -5 ample: A product that performs only automated campaign management *any companies decide to choose point solutions for discrete (2* functions to select the best4of4breed tool for each function <rather than relying on a (2* suite from a single vendor?
Re"erential personali3ation&
0sing e5plicit customer data6such as survey responses, service reAuests, and satis faction feedback6to determine the best selling strategies and marketing messages for a customer
Relationship marketing&
Retention&
(ompany,s ability to keep customers by offering products and services6and, by e5tension, the right messages6to keep the customers satisfied and avoid losing them to a competitor
Rules repositor!&
7ile or database con taining business rules to ensure they are tracked and understood -5ample: -ach time a customer opens an account online, e4mail him a welcome letter
Rules-based personali3ation&
!nvolves the coding of user4defined rules that are analy'ed by the specific personali'ation software tool and used to create tailored customer messages
+creen pop&
3mall window that appears on a user,s workstation screen to provide conte5tual information (ould be a survey <B+ow do you like this Web site so farMC? or a response to a marketing analyst,s mouse click on a certain product, listing the product,s specifications
+cripting&
Automatic BscriptsC <prepared Auestions or comments? generated for customer service reps based on an individual customer,s segment andTor customer profile contents 3cripts remove the guesswork from determining how to respond to a customer Auery or complaint, guiding reps through a dialog with the customer and thus optimi'ing discrete customer interactions
+egmentation&
+el"-ser9ice&
(ustomers can ask their own Auestions or resolve problems without the intervention of a live person 0sually performed over the Web
+kunkworks CR#&
Pro#ect being developed Bunder the radar,C apart from the sanctioned (2* pro#ect that might be underway 0sually the result of unhappy political situations, skunkworks pro#ects can #eopardi'e the reputation and funding of the authori'ed (2* pro#ect by raising Auestions about its value, time liness, or planned deliverables
+teering committee&
8roup of managers or e5ecutives charged with planning and prioriti'ing (2* functionality and allocating appropriate funds for (2* development, rollout, training, and usage (an also influence (2* adoption rates and accompanying business process changes
+tick!&
Web site that grabs and keeps a visitor /ften #u5taposed with the term Beyeballs,C implying that one,s eyes <figuratively? stick to an effective Web site
+uite&
2ange of functional software modules that interact with each other /ften combines marketing, sales force automation, and analytical functions Advantage: (2* suites eliminate integration comple5ity
Target marketing&
Dividing the sum of the customer base into discrete subsets that range from large <dividing customers based on whether or not they own a product? to small <even indi vidual Bsegments of oneC?
Touchpoint&
Point of interaction when the company communicates with a customer, or vice versa /ne interaction6a customer order, for instance6can involve several touchpoints: comparing products on the Web <touchpoint F?, checking inventory levels <touchpoint =?, and placing the order with the sales representative <touchpoint I?
Trouble ticket&
2ecord of a customer,s call into the call center 0sually contains identifying features such as the reason for the customer,s call, the status of the problem, and the ultimate resolution of the call "rouble tickets enable a company to track and monitor the superset of customer calls into the contact center to summari'e the main reasons for inbound customer contacts, whether problems, Auestions, or service reAuests
8p-selling&
*otivating a customer to trade up to a more e5pensive or profitable product "he logic is, now that ! know what this customer wants to buy, perhaps we can motivate him to buy a more profitable version or model -5ample: A #eweler might convince the buyer of a diamond tennis bracelet to go for larger diamonds, at a higher price
=ertical silo&
An organi'ation or system involving a specific and often narrow business function -5ample: A financial analysis system specific to a certain car model and not the automaker,s entire line "he opposite of a vertical silo is a cro""#functional system
=iral marketing&
Word4of4mouth on steroids (ould be positive or negative, but is definitely effective
5ater"all de9elopment&
*ethodical and linear approach to technical development Although rigorous, usually implies going through an entire development life cycle before deploy ing functionality to end users Possible only when e5haustive, detailed reAuirements are known up4front) has been known to take years "he opposite of the waterfall approach is iterative or incremental development, the ideal means of deploying regular (2* functionality
5ork"orce management&
0sually applied to staffing the customer support center Deals with optimi'ing staffing levels in terms of both numbers and skill sets Workforce management tools can analy'e historical call types and volumes and help suggest optimal call center staffing