Test Marketing of New Consumer Products

Author(s): Frank Ladik, Leonard Kent and Perham C. Nahl

Source: Journal of Marketing, Vol. 24, No. 4 (Apr., 1960), pp. 29-34
Published by: American Marketing Association
Stable URL: http://www.jstor.org/stable/1248401
Accessed: 20-04-2015 10:14 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact support@jstor.org.

American Marketing Association is collaborating with JSTOR to digitize, preserve and extend access to Journal of Marketing.

http://www.jstor.org

This content downloaded from 193.226.34.226 on Mon, 20 Apr 2015 10:14:44 UTC
All use subject to JSTOR Terms and Conditions

of
Marketing
Consumer
Products

Test

New

* FRANKLADIK,LEONARDKENT,and PERHAMC. NAHL

Why should new consumer products be tried out in test markets?
Thisarticle points out things to do, and things to avoid.
The design, control, and evaluation of test-marketing programs are
outlined in terms of proper selection of markets, adequacy of the design for measuring results, and optimum length of tests. Number of
markets, measurementsneeded at each stage of testing, and cost and
timing considerationsare discussed.
PROBABLY the most important purpose

of test-marketing programs is to aid in

evaluating opportunities for new products.
Of course, the term "new" may have a
number of different meanings. Some new
products never have been offered to consumers, while others represent simply a
change in form, content, or perhaps packaging. Sometimes the term "new" has been
used when the only change has been one
in marketing approach. Whatever the precise meaning, in each case it has offered a
challenge to the marketing team, including advertising, to do a more effective selling job.
Because of increased emphasis on newproduct programs and the corollary em* About the Authors. Frank Ladik is a Research Supervisor with Needham, Louis and Brorby advertising
agency, Chicago. Prior to joining NL&B, and after
graduating from the University of Illinois with a
B.S. in Marketing, Mr. Ladik spent several years with
a major food processor as an economic and market
research analyst, with extensive experience in newproduct programs.
Leonard Kent is Vice President and Director of
Research of Needham, Louis and Brorby. Dr. Kent
received his Ph.D. in Statistics and Economics at the
University of Chicago, where he taught for eight
years. He had been a member of the firm of Alderson and Sessions in Philadelphia doing marketing
research and management counseling.
Perham C. Nahl is Director of Creative Research
for Needham, Louis and Brorby. Dr. Nahl received
his Ph.D. in Economics at the University of California, and taught there, at Oklahoma A. and M.
College, and at Northwestern University. During
World War II he headed the planning division for
the rationing of shoes and rubber footwear in the
OPA, Washington.

phasis on test marketing, a set of general

procedures and limitations is presented
in this article. These can be of assistance
in the design, control, and evaluation of
test-marketing programs.
Each test-marketing situation is unique.
Each client, brand, product, and market
situation will make somewhat different
demands on a test-marketing program.
Therefore, the intent is to cover only the
more basic principles of test marketing,
recognizing that these principles cannot
always be strictly adhered to in actual operations.
TEST MARKETING
IN INTRODUCING
A NEW PRODUCT

In any organized new-product program,

there is a screening process which begins
with the elimination and selection of potential products or variations at the "idea"
stage and continues through the various
stages of development until finally the
"new" product is considered ready for
sale. More and more companies are reducing the risks of faulty marketing decisions
by first offering the product on a limited
basis.
It is axiomatic that product acceptability should be thoroughly tested through
the laboratory and through consumer research before the product is offered
for sale. Consumer research can indicate
whether or not consumers "like" the new
product, can pinpoint some of their likes
and dislikes, and can provide comparisons
of the new product with those already
29

This content downloaded from 193.226.34.226 on Mon, 20 Apr 2015 10:14:44 UTC
All use subject to JSTOR Terms and Conditions

30

JOURNALOF MARKETING

being sold. However, the only way a manufacturer or distributor can really know
whether consumers will buy the product is
to offer it for sale. If done on a limited
basis, the marketer stands to benefit from
the test program, even if the introduction
is not successful.
A successful operation not only provides a "go ahead" for broadening the introduction, but also sound information
for planning and launching the expanded
program. It is not enough to know whether
consumers will buy the brand. It is also
important to know how much the consumer will buy, how often, and what
profit this volume will yield.
If the test operation shows negative results, the marketer conserves his capital
and other resources with a minimum adverse effect on the company's reputation,
and gains information and experience
that will prove valuable later.
Test marketing, then, involves offering
the product for sale in a limited geographic area (or areas), resembling as
nearly as possible the expanded market
in which the product eventually may be
sold. A program of test marketing must
be carried out with a methodology and on
a scale which will permit reasonably accurate projection of its results.
Such tests should be conducted when
the potential financial loss to the company
through new-product failure, or through
the failure of a market plan or approach,
is significantly greater than the cost of the
proposed test-marketing procedure. This
assumes not only that the new product is
potentially important to the manufacturer
or distributor, but also that any variables
to be tested are of such significance that
measurable variations in results may be
expected.
WHAT CAN BE LEARNED
THROUGHTEST MARKETING?

Answers provided by test marketing obviously will vary as test objectives vary.
However, market tests can be designed to
provide the following different types of
information:

April 1960

maximum profitability. It is not enough to

establish that a previously decided upon
minimum share of the market can be

attained with a specific level of selling,

2.

3.

4.

5.

promotional, and advertising expenditure. Effective test marketing should help

establish the optimum combination of
volume, marketing expenditure, and
profit. While sales goals and expenditure
levels frequently are set on the basis of
judgment, it is usually impossible to forecast accurately the share or volume potentials of a new product.
Test marketing provides an opportunity
for determining the market profile of a
new product. This includes characteristics
of consumers, the way the product is
used, frequency of use, and purchase
history . . . including frequency of purchase, who buys, who makes buying decisions, etc.
Test-marketing performance offers an
opportunity to evaluate the effectiveness
of a marketing program (or of alternative
marketing programs), particularly in the
areas of securing trial and usage of the
product and evaluating amounts expended for advertising and promotion.
While copy testing along the lines of
believability, understanding, and copypoint registration normally is completed
prior to test marketing, the test operation provides an opportunity for further
checking of these factors.
Test marketing can provide up-to-date
trade information which contributes to
the evaluation of the test-market opera-

tion throughthe use of specialaudits and

surveys. Information, both for the test

product and for competitive products,
can be gathered on trade attitudes, trade
allowances and discounts, price-off deals,
displays, spot-advertising expenditures,
promotions, distribution and out of
stock, inventory situations, shelf facings
and positioning, and sales-force performance and requirements.
6. Test marketing offers an opportunity to
obtain from consumers a wide variety of
information relevant to the test operation, including facts, opinions, and attitudes concerning the test product and
competitive products.
HOW MUCH CAN BE TESTED?

There is a practical limit to the numi. Rather than merely showing whether a
product can be sold at a profit, test mar- ber of market programs, or variations,
keting can and should be used to indicate which can be tested at one time. This

This content downloaded from 193.226.34.226 on Mon, 20 Apr 2015 10:14:44 UTC
All use subject to JSTOR Terms and Conditions

31

OF NEW CONSUMERPRODUCTS
TEST MARKETING

limit is for the most part economic. From

a research point of view, it is possible to
test almost any number of variables or
groups of variables, including such things
as different price levels, sizes, media combinations, advertising themes, and consumer and dealer promotions.
However, principally because of cost
considerations, an expanded program to
cover many of these variables is not practicable. A minimum number of test variations will hold losses to a minimum if the
product or marketing variations prove unsuccessful.
On the other hand, should the product
or marketing variations prove successful,
additional information probably will be
desired and required. This additional information usually can be determined more
efficiently (faster and at less cost) during
the conduct of the original test series.
QUALIFICATIONSCONCERNING
TEST MARKETING

If test marketing is to be conducted,

there are certain conditions which must
be met before test results may be used
with confidence:
The objectives of the test must be specific and well defined. The test must be
well designed. It must be well supervised
and analyzed.
The following points also need to be
considered for a program of test marketing:
i. Product or program variations must be
large enough to provide for measurable
differencesin results within the budget
and time requirements.
2. When seeking advertising and promotional expenditure levels, it is desirable
to test at least two variations-a maximum and a minimum.All too frequently,
an expenditure level for testing is
selected on a rather arbitraryestimate of
what is affordable,based on anticipated
sales and earnings. Typically, estimates

same time (although different markets

may be required).
4. If a new variation is to be tested, and

the results compared with those of a

previous test of another variation, it is

desirable that both be tested concurrently. This is to eliminate possible
differences due to the time factor.
5. All factors other than those being tested

should be held as constantas practicable,

and allowed for in evaluating results.

6. Criteria for action should be established,
at least in general terms, prior to testing.
Often market findings during the test
period will indicate a need for modifying
previously established criteria.
HOW LONG IS THE OPTIMUM
TESTINGPERIOD?

One of the most serious and potentially

expensive errors facing the test-marketing
program is the temptation to utilize short
cuts. And one of the more frequent and
dangerous of these is the cutting short of
the test run.
Test marketing has been defined as a

method of reducing future marketing

risks, and thus represents a form of marketing insurance. This insurance, however, is not without its costs. These costs,
or premiums, take many forms, including

the expenditure of money, time, and effort in carrying out an introductory marketing program; the exposure of a new
product and its marketing program for
all competitors to see and analyze; and
commitments

to the trade.

All of these are "premiums" paid for

this form of marketing insurance. To reduce seriously the length of the test run
is to pay these premiums in full without

realizing the full insurance protection.

The actual length of the test-marketing
program should vary, of course, according
to the type of product, consumer buying
frequency, the nature of the industry and

competition, and the activities being

tested. However, there are some general

of sales and earningsare conservative;for rules:

this reason many productsnever have an
i. In introducingnew products,particularly
in the grocery-products field, it is best
opportunity to show their true potential.
Only through testing more than one exgenerally if test periods run at least six
months. This much time is usually
penditure level can a reasonableestimate
of a product's potential be determined.
necessary to evaluate brand progress
3. All important variations to be tested
properly. However, if dealing or couponshould be tested, if practicable, at the
ing continues throughmuch of the intro

This content downloaded from 193.226.34.226 on Mon, 20 Apr 2015 10:14:44 UTC
All use subject to JSTOR Terms and Conditions

32

JOURNALOF MARKETING

ductory period, then in all probability

the test period should be extended to
allow a "normal" sales level to come
about.
2. If there is to be an evaluation of advertising campaignsin connection with the
market test, the concept of cumulative
effect through repetition requires that
the test be run considerablylonger than
the usual product-acceptancetest. It may
well be necessaryto run the test as long
as a year.

3. For promotional tests, the governing time

factor should be the time required for

consumers to return to normal buying
patterns after their exposure to "deal"
merchandise.

Of course, there are substantial pressures exerted-both within and outside

the company-as the test progresses, which
agitate for a decision before a decision is
justified.
Ifnracompany Pressures

The drive and competitive spirit which

contribute to successful management and
aggressive sales policies frequently are
characterized by a restlessness and an
eagerness to "get going." This restlessness
may not only be responsible for setting
test machinery in motion before it has
been adequately prepared, but may also
be responsible for premature interpretation of test data.
Another pressure, perhaps one of the
most difficult to combat, results primarily
from an unwillingness to accept negative
test findings. This lack of objectivity is
understandable when it is remembered
that by the time the product has been in
the test-market stage for some time, management probably has made a considerable investment in ideas, talent, time,
money, and facilities.
Still another form of intracompany pressure comes from members of the sales
force not participating in the test. For any
number of reasons, including rumors of
test-market successes, this group may actively solicit, and in some cases demand,
premature expansion of the marketing
program.
The carrying out of a reliable test operation rests in no small part on the successful resistance to such pressures.

April 1960

Market Pressures

There are numerous market pressures,

including reputed advantages of market
leadership and competitive challenge for
market position. Sometimes there is opportunity in the face of weak demand for
present products. Others are consumers'
loyalty to established brands, and the
"bandwagon."
Management, in its earnest desire to be
first, or its desire to protect brand position
and franchise, or in the face of a slackening demand for its regular products, is
sometimes so impatient that it will rely
upon preliminary test results. If management believes that competitors will respond quickly and that consumer loyalty
to any one brand is a fleeting thing, the
temptation is to expand the marketing
program beyond the test stage before the
facts warrant this.
Perhaps the most dangerous form of
pressure from the outside comes from
trade sources. Wholesalers and dealers,
not a part of the original testing program,
may hear about a particularly promising
new product and exert great pressure on
management to "get on the bandwagon."
These pressures may be hard to resist, especially where excess production capacity
or excess sales capacity exists.
Since even the most expertly conducted
test-marketing programs cannot prove
conclusively that a product will be a success, it is not difficult to understand management's willingness to go along with
the various pressures exerted upon it. The
best that can be hoped for is that, after
committing substantial sums to testing
and paying the premiums for marketing
insurance, management will view the pressures as objectively as possible.
HOW MANY TEST MARKETSARE REQUIRED?

There is no fixed answer to the question of how many test markets are required in a test-marketing operation. As
the number of test markets for a given
variable increases, accuracy increases. And,
as the number of markets increases, the
cost of testing increases. Obviously a compromise must be reached between the accuracy required in the test and the

This content downloaded from 193.226.34.226 on Mon, 20 Apr 2015 10:14:44 UTC
All use subject to JSTOR Terms and Conditions

TEST MARKETINGOF NEW CONSUMERPRODUCTS

amount of money to be appropriated for

the testing operation.
Disregarding the matter of money available, there are several basic considerations
which help to determine the minimum
number of markets to be considered:
1. There should not be fewer than two
markets for each variation to be tested.
This does not include control marketsmarkets in which the variables to be
tested are held constant or allowed for.
2. Where the purpose of the test is to
estimate the sales potential of a product
which is to be distributed nationally,
marketsin at least four geographicareas
should be used.
3. As the significanceof the variablesto be
tested decreases,the number of markets
necessary to reflect the effect of these
variables increases. Thus, it is likely
that where variations are minor, it may
not pay to test market them. The determination of which of two secondary
copy phrases is more effective in building volume is an example of the type of
variable it probably would not pay to
markettest.
These considerations, relating to the
number of markets required, represent an
ideal. In practice, the economics of the
test situation usually compels a reduction
in the number of markets. The extent of
this reduction is, quite naturally, a managerial decision. Care should be taken,
however, that management does not pare
the number of markets to be used to a
point where confidence levels may not be
reliable.

33

3. The market should represent a typical

competitiveadvertisingsituation.
4. The marketshould not be dominatedby
one industry.
5. State capitals, highly industrializedareas
(whereshut-downscould seriouslyimpair
buying power), college towns, and other
areas where population characteristics
not normalto the product'stargetmarket
usuallyshould be avoided.
6. Projectabilityof results is an important
factor. Therefore, the markets selected
should represent different geographic
regions where varying conditions of use
might influencesales.
7. The marketsselectedshould be relatively
independent, with little strong outside
media competitionand relativelylittle of
the test market media circulation going
outside the areas.The same principle applies to the movement of the product
classbeing tested.
8. The markets should have a media pattern which conforms closely to the proposed national media plan. For example,
TV set saturationshould be close to the
U. S. average.
9. Marketsselected should not be too small
to provide meaningful results nor so
large that testing becomes too expensive
in termsof the resultsexpected.
WHAT SORT OF MEASUREMENTS
ARE NECESSARY?

Specific research methods and their

timing will vary with the objectives of the
test-marketing program. This again emphasizes the importance of establishing
precise test objectives in advance of the
test operation.
Fundamentally there are two types of
CRITERIAFOR SELECTINGTEST MARKETS measurements made in market-testing opBecause of the necessity for projecta- erations: (i) consumer studies involving
bility and comparability of test-market usage, habits, reaction, etc., to the brand
results, markets selected for use in the test- in question; and (2) sales measurements
ing operation should meet certain require- involving the brand in question, other
ments. Depending upon the purpose of "company" brands which might be afthe test, not all of these may receive equal fected, and selected competitive brands or
weight, but the list which follows includes products.
When measuring consumer usage, habmost of the generally accepted criteria:
its, and reactions to the brand, interview1. The market or markets should not be
is usually conducted in four stages,
over-tested.On the other hand, theremay ing
which
may, of course, be combined.
be instanceswhen using a marketwith a
"trackrecord"should be considered.
i. The "base" stage-interviewing during
2. The market should have normal historithis stage takes place immediatelybefore
cal development in the product class.
the primarydistributionphaseof the test-

This content downloaded from 193.226.34.226 on Mon, 20 Apr 2015 10:14:44 UTC
All use subject to JSTOR Terms and Conditions

34

JOURNALOF MARKETING

marketing program. Its purpose is to

establish reference points for use in
future analysis.
2. The initial trial stage-interviewing during this stage takes place fairly shortly
after the beginning of the test. Its primary purpose is to determine preliminary
levels of trial attained.
3. Preliminary usage and reaction stageinterviewing during this stage takes place
periodically during the progress of the
test. Its purpose is to determine early
usage levels and reactions to the product.
This stage may phase into the final stage.
4. Final stage-interviewing
during this
stage frequently takes place about six
months after the test has been in
progress. Its purpose is to determine
current usage levels and reaction to the
product after continued exposure.

April 1960

tions in the number of business days from

month to month. If store-audit information is to be obtained, audits should begin
at least four weeks prior to the beginning
of the market test in order to establish a
base for measuring the product's impact
on its competitive environment.
BENEFITSOF TEST MARKETING

No practical program of test marketing

can predict with absolute certainty the
success or failure of products in the market place. Nevertheless, properly planned
and executed test-marketing programs do
represent an important means of examining the growth and profit potentials of a
company's products.
The benefits are worthwhile either when
the product survives, or when the results
The in-home interview is probably the are negative. Survival should contribute
most satisfactory method of obtaining the to greater
profitability of the expanded
above information. However, telephone operation
through information and exinterviews or mail questionnaires may be perience
gained under actual marketing
used with relative success.
operations. In addition to enabling manSales measurements for most products
agement to proceed with greater confiare satisfactory only if made on a retail- dence and
possibly to effect indicated restore audit basis. Shipment data are satis- visions, a successful test
provides evidence
factory only for long-term analysis because to help "sell" the company's own organof distortions due to filling pipelines and ization and trade on the
product.
fluctuations in inventory levels. And shipWhen a test has negative results, this is
ment information does not determine the a small
price to pay for a sound investeffect on competition.
ment. There are, of course, instances where
Store audits at four-week intervals are test
marketing is not feasible; and in these
generally the most satisfactory sales meas- cases judgment must prevail, even as it
urements at the retail level; monthly does in many
phases of a sound test-maraudits are less desirable because of varia- ket program.

MARKETINGMEMO
Marketingas a Science . . .
Marketing as a science . . . emerged as a separate discipline in the

twentieth century and was oriented to immediate results in the business world. Perhapsits greatestweakness,until recently, has been that
it had concentratedon the attempt to establishcontrol without
giving
adequate attention to the companion steps of explanation and prediction. This has tended to make marketingresearchan opportunistic
approachto concreteproblems.
-Wroe Alderson
Marketing Behavior and Executive Action

(Homewood,Illinois:

Richard D. Irwin, Inc., 1957), p. 9.

This content downloaded from 193.226.34.226 on Mon, 20 Apr 2015 10:14:44 UTC
All use subject to JSTOR Terms and Conditions