You are on page 1of 56

Agile Project Management

UB-AMRITA MS IN MITES PROGRAM


ANALYSIS, MODELING, AND DESIGN MODULE
JULY/AUGUST 2007

DELIVERED BY
PROFESSOR RAJIV KISHORE
SCHOOL OF MANAGEMENT
SUNY AT BUFFALO, USA
AMD.MITES@GMAIL.COM

Professor Kishore, 2007 1


Agenda
2

The Mistakes We Continue to Make!


Agile Philosophy and Principles
Jim Highsmith’s Agile Project Management (APM)
Model
Some Prominent Agile Methods
Agile Success Stories
Issues and Challenges
Discussion

Professor Kishore, 2007 2


IT Project Success Rates
3

Failed
15%

Successful
34%

Success Factors:
1. On-time
2. Within budget
3. Needs Met
Challenged
51%

(2004 Standish Group International, Inc. Data


(http://www.softwaremag.com/L.cfm?Doc=newsletter/2004-01-15/Standish)

Professor Kishore, 2007 3


The Complete Software Systems Life Cycle
4

Project Phase Post-Project Use Phase


Software Systems Software Systems
Development Use
Project Post-Project
Failures Failures
The generally But the roots for
talked about these system
category of failures lie in the
system failures project phase

Professor Kishore, 2007 4


Bad News at a Large Scale…

 “Since 1986, the Internal Revenue Service (IRS) has invested $2.5
billion in Tax Systems Modernization (TSM)… TSM is the
centerpiece of IRS’s vision of virtually paperless tax processing to optimize
operations and serve taxpayers better.”
 “As a critical information systems project that is vulnerable to schedule
delays, cost over-runs, and failure to meet mission goals, in February
1995, TSM was added to GAO’s list of high-risk areas.”
 “Nevertheless, the government’s investment of what could be more
than $8 billion and IRS’s efforts to modernize tax processing are at
serious risk due to remaining pervasive management and technical
weaknesses that impede modernization efforts.”
 U.S. General Accounting Office, Tax Systems Modernization: Management and
Technical Weaknesses Must Be Corrected If Modernization Is to Succeed,
GAO/AIMD-95-156 (Washington, D.C.: July 26, 1995),

Professor Kishore, 2007 5


Bad News Doesn’t Stop…

“The FBI has effectively abandoned its


custom-built Internet surveillance
technology, once known as Carnivore, designed
to read e-mails and other online communications
among suspected criminals, terrorists and spies,
according to bureau oversight reports submitted to
Congress.”
 AP, 01/19/2005
(http://www.cbsnews.com/stories/2005/01/19/tech/main667
705.shtml)

Professor Kishore, 2007 6


Just Continues …

“The FBI said last week it was sending back to the


drawing board its $170 million computer
overhaul, which was intended to give agents and
analysts an instantaneous and paperless way to
manage criminal and terrorism cases.”
 AP, 01/19/2005
(http://www.cbsnews.com/stories/2005/01/19/tech/main667
705.shtml)

Professor Kishore, 2007 7


Lest you think it is just the Government…
8

 Digital Equipment Corporation (1980s)


 The Classic XCON System Failure; sales people resented using the
expert application system for configuring client-ordered computer
systems
• Usefulness and ease of use: field study evidence regarding task
considerations, Mark Keil, Peggy M. Beranek and Benn R. Konsynski,
Decision Support Systems, Volume 13, Issue 1, January 1995, Pages 75-91
 Providian Trust (1995)
 Implementation of Access Plus, a Trust and Custody Management
Software, fails during implementation and use; massive resistance by
Personal Trust Managers
• http://www.hbsp.harvard.edu/b01/en/search/searchResults.jhtml?
Ntx=mode%2Bmatchallpartial&Ntt=providian&N=0&Ntk=main_search

Professor Kishore, 2007 8


More Private Sector IS Failures…
9

Foremostco (2000)
 Inventory and Logistics system failures during use
seriously affect logistics and inventory management of
perishable plant and nursery inventory
• http://www.hbsp.harvard.edu/b01/en/search/searchResults.jhtml?
sid=AKL24WNI0PYCQAKRG5DCELQBKE12GISW&Ntx=mode
%2Bmatchallpartial&Ntt=foremostco&N=0&Ntk=main_search
AT&T Wireless (2003)
 Botched CRM system upgrade, system crashes during
customer number porting
• http://www.cio.com/archive/041504/wireless.html

Professor Kishore, 2007 9


Continuing…
10

Global Telecom Company (2005)


 Perfect technological implementation of CRM system, but
system misaligned with users’ needs and goals; system
“misused” and “gamed” by the salespeople and failed
• http://www.cio.com/archive/071505/leadership.html
...

Professor Kishore, 2007 10


11

WHY?

Professor Kishore, 2007 11


IT Project Failures:
Top Project Risk Factors
12

(From Keil et al., CACM, November 1998)

Professor Kishore, 2007 12


Information Systems Success Model

System
Use
Quality
Individual Organizational
Impact Impact
Information User
Quality Satisfaction

Professor Kishore, 2007 13


Technology Acceptance Model

Usefulness

System
Factors Actual
Intention System
(e.g.,
to Use Use
System
Quality

Ease of Use

Professor Kishore, 2007 14


15

A New Solution:

Agile Methods

Professor Kishore, 2007 15


Fundamental Goals in the Agile Movement
16

Delivering innovative products to customers,


particularly in highly uncertain situations
Creating working environments in which people
look forward to coming to work each day

Professor Kishore, 2007 16


The Agile Manifesto
17

 “We are uncovering better ways of developing products by


doing it and helping others do it. Through this work we
have come to value:
 Individuals and interactions over processes and tools
 Working products over comprehensive documentation
 Customer collaboration over contract negotiations
 Responding to change over following a plan
 That is while there is value in the items on the right, we
value items on the left more”
• http://agilemanifesto.org/

Professor Kishore, 2007 17


Individuals and Interactions
18

It is ultimately unique, talented, and skilled


individuals – individually and collectively – that
build new products and services
Processes provide guidance and support and tools
improve efficiency; however, they cannot replace
people who develop products
No process – not even the most rigorously defined
and followed process – can make up for mediocre
individual capabilities

Professor Kishore, 2007 18


Working Products
19

 Large, front-loaded projects prone to failure because their


teams proceed in a linear fashion without reliable feedback
from customers
 Requirements document verifies that a team has
successfully gathered a set of requirements
 Delivering a set of working product verifies that the
development team can actually deliver something tangible
to the customer
 Agile methods stress the delivery of versions of actual
products
 Working features provide dependable feedback that
requirements documents cannot

Professor Kishore, 2007 19


Customer Collaboration
20

In highly volatile, ambiguous, and uncertain NPD


efforts, a collaborative customer-developer
relationship is key to success
Customers define value – they are the ones who
use the products either for end use or to create
further business value
Don’t confuse customers with other stakeholders
(e.g., project sponsors, regulatory agencies, etc.);
they define constraints whereas customers define
value

Professor Kishore, 2007 20


Responding to Change
21

This is characterized by
 Envision-Explore versus Plan-Do
 Exploration versus Production
 Adapting versus Anticipating
Many IT projects turn into disasters because
companies adopt the mantra “Plan the Work and
Work the Plan”

Professor Kishore, 2007 21


22

The Agile Manifesto


Reexamined

Professor Kishore, 2007 22


The Evolution of Methodologies
23
 All methodologies are based on beliefs about what is at the
organizational core
 In the 1970s and 80s, it was Process – so Process-oriented
methodologies
 In the 1980s and early 90s, it was Data and Information – so
Information Engineering type methodologies
 In the 1990s, the core became Objects that encapsulate both data and
processes – so object-oriented methodologies

 In the late 1990s and 2000s?

 People – this is arguably the most important organizational asset and


truly at the core of any organization! - Enter Agile Methods!!

Professor Kishore, 2007 23


The Evolution of Methodologies (contd.)
24

 Many companies in India have been very quick to reach


CMM level 5 – WHY?
 Smart, innovative, talented, and determined people can achieve
almost any results
 Other methods (e.g., AIML) that focus on people, their
roles, their interactions, and their knowledge, etc. are also
being developed and tested
 Enterprise integration using the agent paradigm: foundations of
multi-agent-based integrative business information systems; by
Rajiv Kishore, Hong Zhang and R. Ramesh, Decision Support
Systems, 42(1), 2006.
 Agile Integration Modeling Language (AIML): A Conceptual
Modeling Grammar for Agile Integrative Business Information
Systems; by Hong Zhang, Rajiv Kishore, Raj Sharman, and Ram
Ramesh, Decision Support Systems, forthcoming.

Professor Kishore, 2007 24


Moving from Anticipation to Adaptation
25

 Traditional product development methods are based on a


process of anticipation
 Define
 Design
 Build
 Agile methods are based on a process of adaptation
 Envision
 Explore
 Adapt
 Very similar to the biological evolution process
 Experimentation (mutation and recombination)
 Exploration (survival of the fittest)
 Refinement (producing more of the survivors)

Professor Kishore, 2007 25


Anticipation versus Adaptation
26

 Anticipatory (or Newtonian) approaches aim to produce


predictable results
 The CMM model is built on this approach
 The issue: predictable result may not be an innovative emergent
result
 Adaptive approaches (based on Complex Adaptive
Systems theory) create emergent results that are
innovative
 Agile methods are built on this approach
 One of the overarching goal of agile methods is to deliver
innovative products that deliver outstanding customer value both
today and tomorrow
 creating innovation is creating something new that we can’t fully
anticipate; it is an emergent result based on a vision

Professor Kishore, 2007 26


Why the Pressure for Adaptation?
27

Supply Push
Falling costs of experimentation
 New tools and technologies available that can enable
implementing the agile principles
Demand Pull
 Need for innovative products
 Need to deliver what customers want at the time of
shipment, which is usually not the same as what the
development team “guessed” during planning

Professor Kishore, 2007 27


28

Agile Principles

Professor Kishore, 2007 28


Agile Guiding Principles
29

Deliver
Customer Value
Champion
Employ Iterative
Technical
Feature Delivery
Excellence

Product Delivery

From Jim Highsmith,


Agile Project
Management: Creating
Leadership-Collaboration Innovative Products,
Addison-Wesley, 2004
Build Simplify
Adaptive Teams
Encourage
Exploration

Professor Kishore, 2007 29


Deliver Customer Value
30

 Meet customer expectations


 Remember expectations are not the same as requirements;
requirements are tangible, expectations are intangible
 Delivery rather than plan/control/compliance
 Control focuses on fixing mistakes and explaining discrepancies
when plans are viewed as correct
 But are plans correct? Can innovation be planned?
 Compliance-biased models produce reports for managers,
accountants, government agencies, legal departments, etc., and
delivery of results falls to second place
 Agile methods perform “just necessary” compliance

Professor Kishore, 2007 30


Employ Iterative, Feature-Based Delivery
31

 Iterative development
 Build a partial version of a product and then expand that version
through successive timeboxes followed by reviews and adaptation
 Timeboxes force closure, they force the team to make something
concrete
 Feature-based delivery
 Customers don’t really care about tasks that are to be performed or
are completed; they understand and care about the features that will
be developed and delivered
 They are forced to focus on tasks (behavioral control) rather than on
product features (outcome-based control) in the absence of feature-
based delivery
 Feature-based delivery also induces double-loop learning
 It also results in progressive risk reduction

Professor Kishore, 2007 31


Champion Technical Excellence
32

Project managers must be champions of technical


excellence
If the core purpose of agile methods is to create
innovative products, and if competent individuals
are a key part of delivering those products, then it is
impossible to support a principle that advocates
technical mediocrity

Professor Kishore, 2007 32


Leadership – Collaboration Management
33

You cannot manage men into battle: You manage


things, you lead people
• Admiral Grace Hopper
Management deals with complexity, leadership deals
with change
Project managers have to be both – managers and
leaders

Professor Kishore, 2007 33


Encourage Exploration
34

Exploration is difficult, it causes FUD


Create a safe environment in which people can
voice outlandish ideas, some of which may be off
the wall but some may turn out to be not so
outlandish after all
Encouragement isn’t enough: Put your money
where your mouth is:
 3M used to have a policy to give researchers a percentage
of company time to investigate ideas of their own
 Google gives 10% of company time for exploring new
ideas

Professor Kishore, 2007 34


Build Adaptive (Self-Organizing, Self-Disciplined)
Teams
35

 In a self-organizing team, individuals


 take responsibility for managing their own workload
 shift work among themselves based on need and best fit, and
 participate in team decision making
 Self-organizing teams are not necessarily self-directing
teams
 Self-organizing requires and breeds self-responsibility
 this minimizes control overhead because instead of using process-
based controls, individuals are using self-control
 Self-organizing means you need people who possess a high
degree of self-discipline to begin with

Professor Kishore, 2007 35


Simplify
36

 Simplify the rules and let individual intelligence and


initiative prevail!
 The epitome of simple rules is in Nordstrom’s employee
handbook – a 5’x8’ index card that outlines Nordstrom’s
goal of providing outstanding customer service and lists the
company rules
 “Rule #1: Use your good judgment in all situations. There will be no
additional rules” (Collins and Porras 1994)
 Simple rules can lead to complex behaviors such as
innovation and creativity
 Simple rules force social interaction among team members to
coordinate, resolve problems, and generate new solutions
 Intellectual capital (knowledge and innovation) is generated through
social capital (social interactions)

Professor Kishore, 2007 36


37

Jim Highsmith’s
Agile Project
Management Model

Professor Kishore, 2007 37


The APM Process Framework

Envision

Release
Speculate Explore
Plan

Adaptive
Action Completed
Adapt Features

Feature List
Final Close
Product

Professor Kishore, 2007 38 38


Envision
39

Determine the product vision and project scope, the


project community, and how the team will work
together

Professor Kishore, 2007 39


Speculate
40

Develop a feature-based release, milestone, and


iteration plan to deliver on the vision
Speculating is not reckless risk taking; it is to
conjecture something based on incomplete facts or
information

Professor Kishore, 2007 40


Explore
41

Deliver tested features in a short timeframe,


constantly seeking to reduce the risk and uncertainty
of the project

Professor Kishore, 2007 41


Adapt
42

Review the delivered results, the current situation,


and the team’s performance, and adapt as
necessary
This is not control and correction; they imply one
of two things
initial plans were wrong and they need to be corrected; or
 initial plans were correct but execution was wrong and
that needs to be corrected
Adapt implies modifications or changes to cope
with uncertainties that are inherent in innovation

Professor Kishore, 2007 42


Close
43

Conclude the project, pass along key learnings, and


celebrate!

Professor Kishore, 2007 43


44

Prominent Agile Models

Professor Kishore, 2007 44


Agile Development Models
45

Scrum
Dynamic Systems Development Ecosystems
Crystal Methods
Feature Driven Development
Lean Development
Extreme Programming
Adaptive Software Development

Professor Kishore, 2007 45


Scrum
46

Developed in the 1980's and 90's primarily in OO


development circles as a highly iterative
development methodology
Most well known developers were Ken Schwaber,
Jeff Sutherland, and Mike Beedle
Divides development into thirty day iterations
(called 'sprints')
Focuses on project management aspects and
applies close monitoring and control with daily
Scrum meetings
Much less emphasis on engineering practices

Professor Kishore, 2007 46


Scrum
47

Professor Kishore, 2007 47


48

Agile Success Stories

Professor Kishore, 2007 48


Success Stories
49

Equity One (a publicly traded REIT)


 solution that would allow the geographically dispersed
Equity One staff to generate and retrieve property
operating reports
• http://www.altova.com/cust_Agile_Equity.html
A billion dollar chemical manufacturer
 building an enterprise data warehouse
• http://www.cariboulake.com/stories/stories.html
Large County Government agency
 application with four primary subsystems delivered in
phases: Jail, Court, Supervision, and Administration
• http://www.cariboulake.com/stories/stories.html

Professor Kishore, 2007 49


Success Stories (contd.)
50

 Objective Systems
 migrating existing GUI from Windows to the Linux and Unix
platforms
• http://www.sourcextreme.com/objsys-success.html
 Colubris Networks
 implement an enterprise-critical, product lifecycle management
(PLM) solution
• http://sme.agile.com/customers/pdf/cs-colubris-20041005.pdf
 OnStor
 Product Lifecycle Management solution
• http://sme.agile.com/customers/pdf/cs-onstor-20041005.pdf
 Egg Technology
 Development of “Egg Money Manager” to support its personal
financial services
• http://www.mercury.com/uk/pdf/success/egg_success.pdf

Professor Kishore, 2007 50


51

Issues and Challenges

Professor Kishore, 2007 51


The Pros and Cons of Agile
52
Pros Cons
 System functionality is more in  Schedule and costs are less
line with needs predictable
 Faster release cycles  Systems may be harder to
 More satisfied technical maintain
personnel  Architecture may be less robust
 Higher customer commitment and less stable (although
 Highly suitable for projects proponents don’t agree with
with undefined and/or this
 Works best in small projects
changing requirements
– this is a major issue
 Tremendous user commitment
is required

Professor Kishore, 2007 52


Scalability to Large Projects?
53

 Questionable; issues of coordination


 Coordination is managing interdependence; three types of
interdependence
 Pooled
 can be coordinated simply by standards
 Sequential
 Activities are linked sequentially
 more complex coordination needed
 Can be done using hierarchical coordination (e.g., project managers)
 Reciprocal
 very complex coordination needed
 Intensive technologies and the entire group is needed to achieve desired
outcomes (e.g., hospital ERs)

Professor Kishore, 2007 53


Scalability to Large Projects?
54

CMM and other methodologies use standards and


hierarchies as coordinating tools
Agile uses group meetings as the coordination tool
 Issue – managing large group meetings
 Tools and technologies for large group meetings are not
yet there
 We also don’t understand collaboration and coordination
very well in large geographically distributed teams

Professor Kishore, 2007 54


Acceptance of Frame Breaking Ideas
55

Agile requires frame-breaking


Innovations, especially those that require frame-
breaking, have a slower diffusion process
 Think object-orientation
 Nearly four decades since the ideas were first developed
(Simula 67)!
But the diffusion process for agile is on

Professor Kishore, 2007 55


Sources and References
56

 I have drawn from a number of sources in this presentation


and it is difficult to list all of them here
 Some citations were provided in the presentation
 Jim Highsmith’s book “Agile Project Management: Creating
Innovative Products by Jim Highsmith; Addison-Wesley,
2004” has informed this presentation quite a bit. The APM
model presented today is covered in much more details in
his book
 Another excellent source for learning about agile methods
is the website titled “The New Methodology”
http://www.martinfowler.com/articles/newMethodology.h
tml

Professor Kishore, 2007 56

You might also like