|
|
The four-fold purpose of market research
The manufacturer of consumers' goods can
seldom escape the necessity for careful market research and analysis.
Moreover, the process must be continuous. Research and analysis are as
important through the whole life of a manufacturing enterprise as they
are just prior to the introduction of a new product. The work of the market
research department is never done. Consumer demand is changing constantly.
There is on the one hand a need for the compilation continuously of current
data, most of which can be standardized, and on the other hand a need
for periodic but intensive surveys of special problems, no two of which
are likely to be similar.
The fundamental purpose of market research
is four-fold. Market data is necessary:
- To provide the basic factual information
upon which all marketing policies are formulated
- To provide the basic factual information
upon which specific marketing campaigns may be built
- To provide the data for the analysis
and evaluation of the effectiveness and economy of specific marketing
methods and operations
- To assemble the necessary statistical
data and conservative opinion which make market forecasting possible.
Market research basic to marketing policies
The whole range of marketing policies concerning
the product, prices, trade channels, and dealers and consumers should
be based upon factual information which is both comprehensive in extent
and accurate in individual details. The process of formulating marketing
policies should recognize the fact that consumers' demands are the origins
of economic impulses and that these consumers' demands are dynamic and
changing. Research must be instituted to gather all the essential quantitative
facts with respect to the number, geographic distribution and economic
status of present and potential customers. In addition, qualitative data
concerning the preferences, habits, customs, buying motives, and resistances
or prejudices of the various groups of customers must be collected. Both
types of information are necessary because willingness to buy is just
as important as ability to buy. It is the sheerest folly to adopt identification
systems, price-discount ranges, definite trade channels, arbitrary returned-goods
restrictions, and other policies without the most careful study of the
present marketing situation and of the probable reaction to proposed changes.
Research as a basis for market planning
Market research and analysis should also
be employed to determine the selling methods and mechanisms that are necessary
in order to get the maximum results from marketing efforts. Time and energy
must be devoted to the all-important tasks of planning specific marketing
campaigns and building up the sales program in direct relation to particular
requirements. The wastefulness of "blind" or "hunch"
planning is obvious. Campaigns should not be shot wildly into the air.
They must be aimed directly at the target indicated by a careful pre-analysis
of market conditions. A well considered plan of attack will avoid many
serious errors, such, for example, as the one committed by a large paint
concern. This concern sent an army of personal salesmen into a certain
state solely because that state was listed as "good, prosperous territory,"
ignoring or totally unconscious of the fact that 85 per cent of the dwellings
in the territory were constructed of brick and offered at best only a
limited market. In the same way, advertising must be based upon comprehensive
and accurate market information. Advertising wastefulness in far too large
a number of instances results from increasing the volume of advertising
when a certain amount fails to pull, instead of changing the mediums and
the character of the advertising. Certainly, a 100 per cent job cannot
be accomplished by adding in advertising volume to a 50 per cent "content
of weakness and paucity of definition, and a considerable degree of indirection."
Economy and cost investigations necessary
Entirely apart from the task of gathering
basic information upon which marketing policies and specific marketing
campaigns can be set up, the market research department should devote
considerable time and effort to investigations directed toward economy.
Sales operations should be broken down into their basic elements by a
process somewhat similar to that used in time studies of production. Averages
of sales results must be worked out. Ideal standards of sales operation
must be set up, and accomplishments checked against these standards.
Personal selling activities should be studied
and their results evaluated. Ineffective methods and useless and unnecessary
operations will never be scrapped until their wastefulness is emphatically
demonstrated. All down the line of personal-selling activities—number
of sales calls, prospecting, missionary work, etc.,—methods need
to be studied and evaluated. In the same way, advertising methods require
careful study. The copy, illustrations, typography, space utilized, media
employed and all other factors must be tested periodically to make certain
that they meet the standards set by careful research.
Research contributes forecasting data
Finally, the market research department may
be called upon to devote a part of its time to the collection and compilation
of territorial sales records, per capita consumption by territories, and
other statistical data which may be used as indices in connection with
market forecasting for the purpose of determining how sales may be increased
or with the idea of developing new or neglected markets. After a useful
combination of indices has been worked out and a satisfactory method of
weighting has been discovered by means of which forecasting can be carried
on with reasonable accuracy, standards can be set for the type of basic
information required and the method of collecting and compiling it. Then,
if the market research department is charged with the responsibility of
collecting this information, as it usually is, the task can be reduced
largely to a routine.
Methods of gathering data
For reasons of clarity, it may be well to
discuss the first two purposes of market research as one. The technique
of collecting market data as a basis upon which to formulate marketing
policies, and the technique of gathering data on which to build specific
marketing campaigns are one and the same. They differ, however, to some
extent, from the technique of making economy and cost investigations and
from the method of gathering forecasting information. The type of research
that underlies marketing policies and marketing plans has for its chief
focal points, the product, the market, and competition. Every comprehensive
survey should include a careful investigation of these three factors.
Before we attempt to indicate the important points to be sought under
the headings of the product, the market, and competition, we should consider
the methods to be used in gathering such data. The two most common methods
of gathering information from the field are the circulation of questionnaires
by mail and the use of personal interviews. The questionnaire is fundamental,
in most cases, to each one of these methods of field survey. The personal
interviewer conducts his investigation usually from a carefully prepared
list of questions which in a printed and convenient form he may present
directly to the person to whom he is talking. Occasionally, however, he
may fill out the form after he has withdrawn from the presence of the
person being interviewed.
Advantages of mailed questionnaires
The chief advantages of securing information
by means of questionnaires sent through the mails are the low cost of
the method, the wide territory that can be covered, and the speed with
which a particular survey can be completed. This method is widely used
because it is generally considered to be inexpensive. To send out 10,000
letters, each containing a questionnaire, a letter of request, and a return
stamped envelope will cost a lot, including postage for outgoing and return
envelopes. The same number of interviews would, of course, cost very much
more in money and in time. However, if the returns from a mailed questionnaire
are low and if the distribution of replies is not satisfactory, this method
may prove to be just as costly as that of personal interview. Nevertheless,
in theory at least, the mailed questionnaire can often be made to include
a large number of people and people from a wider range of territory than
would be economical by means of personal interviews. Blanks can be mailed
to thousands of people located in all parts of the country and the answers
will be received in a fraction of the time necessary to accomplish the
same extent of survey by personal investigators.
Disadvantages of mailed questionnaires
The principal disadvantages of the mailed
questionnaire method are the small percentage of questionnaires returned,
the fact that questionnaires are usually answered by selected groups of
people, the uneven distribution of replies, and the large amount of useless
and incomplete data which is frequently returned. The matter of securing
reliable mailing lists is not included as an important difficulty or disadvantage
because the number of lists of all kinds which can be secured from responsible
mailing-list companies is continually on the increase. Telephone directories,
city and county records, and the membership rolls of various organizations
are typical of other name sources. Generally speaking, the relative number
of responses obtained from those to whom questionnaires are mailed is
small. The percentage may range from one or two up to fifty. It will depend
largely upon the type of people to whom the questionnaire is sent, their
interest and intelligence, and the persuasiveness of the letter of request
which accompanies the questionnaire.
Moreover, the replies which are returned
tend to come from a selected group rather than a representative or typical
group. The persons who do reply may be more intelligent, more progressive,
more conscientious, more obliging, or more curious than those who do not
reply. And yet those who do not reply may represent the very type of market
opinion which the investigator needs and wants. Again, the difficulty
of controlling the distribution of replies over various districts is exceedingly
great. A 20 per cent response from the Middle Atlantic States as compared
with 3 per cent from the New England States is of little value in drawing
general conclusions if the two sections are of equal importance. Finally,
many people fill out questionnaires very hastily and with their tongues
in their cheeks. Much useless and inaccurate information and some that
is intentionally misleading is quite likely to appear on many of the returned
questionnaires. The instance of the manufacturer of women's wear who on
a widely circulated questionnaire asked a question concerning the make
of automobile owned is typical. The returned questionnaires reported four
times as many Rolls Royce automobiles as were registered in the United
States and Canada.
Use of mailed questionnaires
Because of these difficulties or disadvantages
the mailed-questionnaire method should not be too frequently depended
upon as a single means of collecting market data. In most cases it should
be supplemented by the method of personal interview. Thus, each method
will serve as a' check on the other. The exclusive use of mailed questionnaires
should be confined chiefly to cases where the returned blanks need not
be particularly balanced from the point of view of location or type of
person replying, and where the investigation is so large in scope as to
preclude the use of investigators except incidentally.
Advantages of personal interviews
On the whole, it is far better research practice
to obtain information in person than it is to attempt to secure representative
replies by mail, even though the method of personal interview may result
in a smaller number of responses. Sometimes even a small number of responses
secured by personal interviews, obtained by adhering to a carefully-prepared
questionnaire, will give highly reliable results. Not only does the personal
interview result in a more accurate and representative distribution of
responses, but it usually provides more detailed and complete information.
For one thing, if the interviews are based on carefully prepared questionnaires,
a comparatively large number of questions may be asked and an unusually
large amount of detailed information may be obtained. On the other hand,
if the responses are to be obtained by mail, the questions must of necessity
be few in number, simple, brief and to the point. Again, a personal investigator
can explain perplexing questions in their different shades of meaning
and thus use a certain amount of salesmanship to persuade those who are
being interviewed to reply accurately and in full. Then, too, in the personal
interview there is less likelihood of securing facetious and intentionally
misleading information than there is in the filling out of a mailed questionnaire.
Disadvantages of personal interviews
The principal disadvantages of the method
of personal interview are the expense of widespread surveys, the length
of time necessary to secure information, and the difficulties of conducting
an interview skilfully to secure absolutely accurate data. If special
investigators must be engaged and trained to make the proposed survey,
costs are certain to be important and often prohibitive. To employ experienced
investigators on full time, to train them carefully in the method of carrying
on the survey, and then to send them out into the field is obviously the
most certain way of securing reliable responses. But it is expensive.
Accordingly, many concerns depend to some
extent on their own salesmen, newspaper reporters engaged specially for
the individual job, and correspondent investigators located in various
cities who will also undertake surveys by the job. Fortunately, the number
of organizations prepared to do this type of work on a commercial basis,
is rapidly increasing in most marketing centres. In spite of these growing
opportunities to shift a part of the task of collecting data to specialized
agencies, extensive surveys are costly and slow in completion. Moreover,
information which the people being interviewed consider confidential may
not be given to the investigator, although it may be accurately supplied
on an unsigned, unidentified mailed questionnaire.
Methods of conducting interviews
Of course, personal interviews may be conducted
without the assistance of a carefully-prepared questionnaire. The investigators
may be given general instructions as to the type of information desired
and then left free to conduct the interview as they see fit. But such
a procedure usually means a wide variation in the way in which questions
are presented to the people being interviewed, and a greater possibility
of error on the part of the investigator in remembering the actual responses
obtained. In general, it is wiser to prepare an interviewing questionnaire
and to instruct the investigators to use it and to follow it very closely.
If there should be any objection to the display of the questionnaire in
the presence of the person interviewed, the investigators should be instructed
to commit it to memory so that all of the points will be covered in each
interview. However, the idea that the persons who are being interviewed
will be less inclined to talk freely if the investigator uses a questionnaire
in their presence has little basis in practical experience.
Quantitative and qualitative data
In considering or planning the technique
for conducting surveys, it is well to keep in mind the distinction between
quantitative and qualitative types of information. Obviously, some surveys
are designed to secure simple facts, others to secure the opinions or
judgments of considerable numbers of people. These two types of inquiry,
of course, may be carried on concurrently in the same survey but the question
of relative emphasis on one type of information as compared with the other
should be carefully determined. The relative emphasis determines to a
large extent the investigating technique that should be followed. The
six essential steps in working out the technique for a survey are as follows:
- Determining the field for investigation
- Preparing the questionnaire
- Collecting the data
- Tabulating the data
- Drawing the conclusions
- Presenting the report.
Determining the field of survey
In the case of investigations which aim to
secure factual data in the main, determining the field for investigation
is a comparatively simple and direct task. The problem is largely one
of numbers. In preparing an investigation which is to collect opinions
and judgments, greater care is necessary in the selection of representative
groups of people from whom information is to be sought. Generally, this
type of information is desired from one or more specific groups of consumers,
retailers, wholesalers, producers or specialists (engineers, doctors,
trade authorities, architects, etc.) located in one or more specific trading
areas or territories. The problem is to seek out small groups comprised
of people who will be able to answer questions accurately and intelligently,
and who are typical of the large groups in the market.
Preparing the questionnaire
The preparation of a questionnaire which
is to be used in the collection of factual data necessitates simple and
short questions calling for "yes" or "no" answers.
'When opinions and judgments are sought, the preparation of the questionnaire
calls for more careful planning. The questions must cover the important
points on which information is desired and they must be so worded as to
secure adequate and correct answers. So far as is possible the questionnaire
must be designed to prevent the person being interviewed from distorting
the truth "due to self-consciousness," from saying, for example,
that a certain product is purchased because of its flavour when actually
it is purchased because it is cheap. Also, there should be some leeway
for expressing and recording unusual reactions, opinions and judgments.
Necessity for specific questions
T here are five definite factors essential
to the preparation of a questionnaire, whether the questionnaire is to
be mailed out for responses or is to be used by investigators in personal
interviews. In the first place, the questions asked should be specific
and definite. In-stead of asking, "Do you serve pancakes in the summer
and, if so; how frequently?" it is better to ask, "How many
times a week do you serve pancakes in the winter? In the summer?"
A general question like, "Do you buy many lemons in the winter time?"
is not nearly so satisfactory as a more definite set-up such as : "Do
you buy as many lemons in winter as in summer? Half as many? A third as
many? A fourth? Any other amount?"
Use of the general question
In the second place, a few questions of a
somewhat indefinite nature should be inserted in the questionnaire "to
keep the way open for obtaining any unusual reactions, thus avoiding deadening
or deceptive standardization." These questions will permit the people
from whom responses are requested to give a free and unbiased response.
Many of these responses yield personal or individualistic "slants"
which may prove very valuable in the interpretation of the collected data.
Questions of this type may very well be inserted just before more specific
questions. For example, the question, "What do you like about the
printer you are now employing?" may precede a more specific question
such as, "What influences you in placing printing orders? Please
check in their order of importance, 1, 2, 3, 4, 5, 6, etc.
Avoiding the leading question
The third essential in preparing a questionnaire
is to phrase the questions in a way to avoid warping or prejudicing the
response. Leading questions must be carefully avoided. A question that
begins with the phrase, "Don't you think?" makes a suggestion
that is not present when the same question is introduced by, "What
do you think?" Incidentally, it is in an effort to suggest a desired
response that the department stores instruct their salespeople to say,
"Will you take this parcel with you?" rather than, "Will
you have it sent ?"
Statistical treatment of responses
Questions should be formulated in a manner
to secure responses that will lend themselves to tabulation and statistical
treatment. The question in Section 17, above, concerning printing illustrates
a type of question designed to produce responses that can be treated statistically.
Any question or set of questions embodied in a questionnaire, in which
the answers are ranked in order of importance yields the type of results
which can be tabulated in some usable statistical form.
Number of questions
Finally, as has already been suggested, the
number of questions and the detail in which they are presented depends
largely upon the method used to obtain the responses. If mailed questionnaires
are to be used, the questions presented must be brief and few in number.
Most people are too busy with other matters to want to fill out long and
detailed questionnaires. They seldom feel obliged to complete and return
a mailed questionnaire if the task is at all formidable. On the other
hand, if the information is to be obtained by personal interviews, the
investigator may use successfully a detailed questionnaire which presents
a fairly large number of questions.
Collecting the responses
If factual information is desired the mailed
questionnaire offers usually the most suitable medium, particularly if
the distribution of replies can be controlled. The objective of the questionnaire
should be to secure a large number of answers cheaply, quickly, and impersonally.
If opinions or judgments are sought, personal interviews give results
that are more thorough-going and accurate. They are essential to any inquiry
that is concerned mainly with securing qualitative information. Of course,
the danger that the response may be coloured or modified through the personal
reaction of the investigators must be minimized. For this reason the questionnaire
should be prepared with special care.
Tabulating the data collected
The tabulation of factual data is a rather
simple and direct process. If the questions that appear in a questionnaire
call for direct answers, such as "yes" or "no," a
recapitulation of results is easily made.
Obviously, these answers could also be tabulated
in percentages and set up in the same general form. When, however, the
survey partakes of the nature of a qualitative inquiry, the work of tabulation
is more complicated. If a question is asked which may be answered in a
number of possible ways, the order of merit method may be used.
Numerous other methods exist for summarizing
and analysing data, but they are not discussed here because many of them
are highly involved, and because it is impossible to lay down any hard
and fast rules of tabulation that will be applicable universally, or even
to any two different surveys. The objective in tabulation should be to
provide for comparisons, to insure accuracy, and to preserve those individualistic
qualitative responses that are particularly significant.
Drawing conclusions from tabulated data
After a careful market survey has been made,
and the results have been properly tabulated and compared, the conclusions
or inferences to be drawn would seem to be largely self-evident. Unfortunately,
such is not always the case. A considerable amount of logical induction
and deduction is usually necessary in order to set up sound conclusions.
Reasoning is highly involved. Indeed, it is very largely a matter of tracing
connections of resemblance and Connections of cause and effect. The opportunities
for going astray in drawing conclusions are many. Some of the most common
errors were listed as follows by W. J. Reilly, of Lennen and Mitchell,
Inc. :
- We study a few cases in a certain class
and then proceed to make general statements about every member of that
class.
- We take a general rule that applies to
all cases in a certain class and make the mistake of applying it to
a case 'which appears to belong, but which actually does not belong,
in that class.
- We are misled into believing that objects
or situations that are alike in some respects are alike in other respects.
- When we know that a certain condition
existed at a certain time, we make the mistake of assuming that the
condition was at some previous time, is now, or will be at some later
time necessarily the same.
- When one is unsuccessful in proving a
proposition, he is inclined to believe that the opposite is true.
Now, since reasoning is so largely a matter
of tracing connections, it can be tested best by asking and honestly answering
certain connective questions. In cases of comparison, where the connection
of resemblance is involved, the following questions are helpful:
- Do the apparent connections express essential
likeness?
- Do the essential likenesses outweigh
the essential differences ?
In cases of causal relationship, two other
questions may be asked:
- Is the apparent cause adequate to produce
the recorded effect?
- Could and did other causes operate to
produce the recorded effect?
Presenting the survey report
The problem of presenting the conclusions
of a survey to the management is extremely important. Surveys will not
automatically correct the conditions that they may disclose. Suggested
solutions must be approved, adopted, and put into operation. Consequently,
the matter of presenting the survey in a logical, clear and altogether
persuasive manner is an absolute necessity. "Time after time it happens
that some ignorant or presumptuous member of a committee or a board of
directors will upset the carefully-thought out plan of a man who knows
the facts, simply because the man with the facts cannot present his facts
readily enough to overcome the opposition."
Generally speaking, the proper procedure is to present the survey report
in written form, and then at a suitable interval to follow this presentation
with oral conferences with interested executives. The form of the written
report may well be indicated by the following headings:
- Purpose of this survey
- Description of the field of this survey
- Description of the methods used in this
survey
- Results of this survey (tables, etc.)
- Conclusions
- Related observations
- Summary
The summary should be short, concise, and
non-technical. It should include the definite conclusions and pertinent
and related observations. Throughout the report, pictures, charts, and
graphs should be used whenever possible.
|
