In essence, management is about decision making. Decision making is invariably surrounded by uncertainties and, therefore, risks. Marketing research is charged with helping to reduce the level of uncertainty with which marketing managers must cope. Marketing research is a form changing activity in that it takes facts and figures and other types of raw data and converts this to information, in which form it is useful to decision makers. The presentation of marketing research in this chapter is less from the perspective of practitioners in this field and more from the stance of one who has to draw on the expertise from these specialists and/or perhaps manage them. Thus, the chapter is not about how to carry out a marketing research exercise rather it deals with the information that can be obtained from these fields of expertise, how to commission studies and how to manage these functions.
The reader will observe that the term used here is that of ‘marketing research’ rather than ‘market research’. This is in recognition of the fact that a great deal of the research effort is typically devoted to identifying marketing problems (e.g. improved distribution systems, packaging development, studies of business trends, long range forecasts of demographics etc.) whereas the narrower term, ‘market research’ is indicative only of research into markets.
The specific objectives of this chapter are to assist the reader gain an understanding of:
The role of both marketing research in marketing management decision making
The content of the briefing that those who commission marketing research must give to the individuals charged with carrying out marketing research
What those who commission marketing research must ensure is included in the proposals submitted by internal or external suppliers of marketing research
The structure of a good research report.
The introduction to the chapter seeks to establish the nature and purpose of marketing research in the context of management decision making. This is followed by an outline of the principal contents of the brief given to those chosen to undertake a given marketing research exercise. The main body of the chapter deals extensively with the essential components of the research proposal that is drafted in response to the research brief. The chapter concludes with a brief overview of the structure of marketing research reports.
The term marketing research can be defined as follows1:
“Marketing research is the systematic and objective search for, and analysis of, information relevant to the identification and solution of any problem in the field of marketing.”
Green & Tull (1978)
The key words in this definition are; systematic, objective and analysis.
Systematic: Marketing research seeks to set about its task in a systematic and objective fashion. This means that a detailed and carefully design research plan is developed in which each stage of the research is specified. Such a research plan is only considered adequate if specifies; the research problem in concise and precise terms, the information necessary to address the problem, the methods to be employed in gathering the information and the analytical techniques to be used to interpret it.
Objectivity: Maintaining objectivity in marketing research is essential if marketing management is to have sufficient confidence in its results to be prepared to take risky decisions based upon those results. To this end, as far as possible, marketing researchers employ the scientific method. The characteristics of the scientific method are that it translates personal prejudices, notions and opinions into explicit propositions (or hypothesis). These are tested empirically. At the same time, alternative explanations of the event or phenomena in which we are interested are given equal consideration.
There is a tendency for marketing managers to be prejudice in favour of a proposed project, such as entry into a new market or the development of a new product, once it has been decided that his/her organisation should formally investigate its potential. This is especially the case where the manager has at some point supported, or even originated the project. He/she may even specify the research problem in a way that biases the results towards a positive answer, e.g. “Your job is to find out how big the market might be for this product”. In reality, the task is to determine whether or not, there is a market for product X. The marketing researcher has to resist the pressures to simply confirm the prejudices of the person who has commissioned the study.
Analytical: The third of the key terms in the definition given a little earlier was analytical. The marketing researcher's task goes beyond the collecting of data. He/she must also interpret it in terms of what the data means to the organisation which commissioned the research. Knowing that sixty percent of those interviewed thought that product A was superior to product B is, in its self, of little value. The organisation needs to know the alternative ways it can respond to this data. Data is equivalent to the raw materials of manufacturing it has to be converted to information before it becomes useful in decision making. The process of converting data into information is achieved through analysis.
Whilst there is a need for accuracy, precision and thoroughness in marketing research it is to be remembered that, in practice, there is a perpetual conflict between the demands of expediency and the search for truth. The reality is that management is frequently under pressure to make timely decisions. Therefore management often seeks answers through marketing research in the shortest time possible and moreover, at minimum cost. On such occasions its methods tend to be less theoretically rigorous and its analysis more superficial.
Marketing research can be concerned with any of a variety of aspects of the market; the product, sales, buyer behaviour, promotion, distribution, pricing, packaging etc. Since the researcher cannot investigate everything about a market, he/she must be selective. The question remains as to how the researcher decides where to focus the study; and to what depth each issue should be investigated. The answers should lie in a document called the research brief. The research brief is a set of guidelines given to the researcher by the person(s) who have commissioned the research and/or the individual(s) who are to make us of the results in their decision making. The brief must inform the researcher which aspects of the market are particularly important. In particular, the research brief should include:
The purpose of the research
The objectives stated in a clear, concise, attainable, measurable and quantifiable way
A time horizon
A resource allocation, including the budget and facilities
A reporting period.
Each of these components of a research brief is expanded upon below.
It is not at all unusual for marketing managers to neglect to tell the researcher the precise purpose of the research. They often do not appreciate the need to do so. Instead, they simply state what they think they need to know. This is not quite the same thing. To appreciate the difference consider the following case.
A marketing research agency was contacted by the International Coffee Organisation (ICO) and asked to carry out a survey of young people in the age group 15–24 years of age. They wanted information on the coffee drinking habits of these young people; how much coffee, they drank, at what times of day, with meals or between meals, instant or ground coffee, which other beverages they preferred and so on. In response, the research organisation developed a set of wide ranging proposals which included taking a large random sample of young people.
In fact, much of the information was interesting rather than important. Important information is that which directly assists in making decisions and the ICO had not told the research company the purpose of the research. The initial reason for the study had been a suspicion, on the part of the ICO, that an increasing percentage of young people were consuming beverages other than coffee -particularly soft drinks- and simply never developed the coffee drinking habit. Had this been explained to the research company then it is likely that their proposals would have been radically different. To begin with, the sample would have been composed of 15–24 year old non-coffee drinkers rather than a random sample of all 15–24 year olds. Second, the focus would have been non-coffee drinking habits rather than coffee drinking habits. Unless the purpose of the research is stated in unambiguous terms it is difficult for the marketing researcher to translate the decision-maker's problem into a research problem and study design.
Suppose that the marketing manager states that he/she needs to know the potential market for a new product his organisation have been developing. At first glance this might appear to meet all of the requirements of being clear, concise, attainable, measurable and quantifiable. In practice, it would possible meet only one of these criteria, i.e. it is concise. The problem with the objective; that the needs to know the potential market for the new product, is that it is not attainable. One could find out how many tree lifters were currently being sold but this is not the same as the objective set by the marketing manager. As figure 11.1 suggests the market potential for any new brand is a function of at least 4 things: customer reaction, competitor reaction, the marketing mix and trends in the environment.
Figure 11.1 Factors influencing demand for a new product
Consider the case of the small West African engineering company that wanted to diversify its product line and had purchased the designs for a cassava grater. Once grated the cassava would be mulched, allowed to ferment for some days and then by progressively drying the mulch at carefully controlled declining temperatures would produce gari, a foodstuff which when reconstituted with hot water resulted in a glutinous porridge. The cassava grater could be constructed in wood or manufactured in sheet steel. The wooden model was designed to be hand driven while the steel model could be powered by means of 1.5 horsepower petrol engine. The company wisely decided to conduct a market study before launching the product but posed the research question as : ‘What is the market potential for the new cassava grater?’
It was possible to test consumer reaction to the concept of the new cassava grater by showing them pictures, line drawings and by supplying product specifications to prospective buyers. However, since the company had not decided their pricing policy an important element could not be tested. In large measure, it was also possible to gauge the likely reaction from competitors. The researchers began by looking at the basis of competition was it on price, product quality or unique product features? The researchers were able to look at precedents. They examined the pattern of response on past occasions when one or other of those companies already in the market had launched a new product. An audit of the environment was undertaken too but the missing component was the companies' own plans for exploiting the market, i.e. its marketing mix. Since the company had no involvement in the agricultural engineering sector prior to acquiring the rights to the cassava grater they had no agreements with distributors, no idea of which, if any, of the distributors would be prepared to stock their product, they had no salesmen trained in selling into this industry and so on. The product's potential depended very much on such initiatives.
The solution would have been to have undertaken a study which would have described the market in detail in terms of customers, competitors and the environment. The company could then have put a marketing plan together and then conducted a follow up study to test their propositions out on the marketplace.
The need to set a time horizon for marketing research
Inevitably there are deadlines which the marketing research activity must fit and these must be stated clearly at the outset of the research. As was said earlier, due to time pressures management is often seeking quick answers from marketing research. If the researcher is aware of the time constraints then this will become an overriding factor when he/she plans the research design. He or she is likely to put forward a design which is less elegant, and gives rise to less precise information but delivers the results on schedule.
Resource allocation, including the budget and facilities
There are essentially two approaches to establishing the resource allocation to a particular marketing research exercise. Researchers can start with the problem and work out how much it will cost to solve it. This figure, along with the research design, can then be put to the person commissioning the work. Alternatively, manager can decide how much can be afforded and seek the best answer obtainable for the time, money and manpower allocated. In practice the decision-makers prefer the latter approach and the researchers the former. In the end, some kind of compromise develops. The researcher rarely gets all of what he/she judges is required to reach a satisfactory conclusion but if the research proposal is well thought out and persuasively presented some concessions can be obtained. Whichever the approach to resource allocation adopted, it is imperative that the researcher is aware of the financial, and other constraints within which he/she must complete the work and also the study the points in time when interim reports are required, if any, and the deadline for the final report.
Having received the research brief, the researcher responds with a research proposal. This is a document which is developed after careful consideration has been given to the contents of the research brief. The research proposal sets out the research design and the procedures proposed be followed. The eight steps are set out in figure 11.2.
Figure 11.2 The research design
The point has already been made that the decision-maker should clearly communicate the purpose of the research to the marketing researcher but it is often the case that the objectives are not fully explained to the individual carrying out the study. Decision-makers seldom work out their objectives fully or, if they have, they are not willing to fully disclose them. In theory, responsibility for ensuring that the research proceeds along clearly defined lines rests with the decision-maker. In many instances the researcher has to take the initiative.
In situations, in which the researcher senses that the decision-maker is either unwilling or unable to fully articulate the objectives then he/she will have to pursue an indirect line of questioning. One approach is to take the problem statement supplied by the decision-maker and to break this down into key components and/or terms and to explore these with the decision-maker. For example, we could ask what he has in mind when he uses the term ‘market potential’. This is a legitimate question since the researcher is charged with the responsibility to develop a research design which will provide the right kind of information. Another approach is to focus the discussions on the decisions which would be made given alternative findings which the study might come up with. This process frequently proves of great value to the decision-maker in that it helps him/her think through the objectives and perhaps select the most important of the objectives.
Whilst seeking to clarify the objectives of the research it is usually worthwhile having discussions with other levels of management who have some understanding of the marketing problem and/or the surrounding issues.
Other helpful procedures include brainstorming, reviews of research on related problems and researching secondary sources of information as well as studying competitive products.
Whilst it is true that the purpose of research is to address some question nonetheless one does not test research questions directly. We may, for example, be interested in answering the question; ‘Does a persons level of education have any bearing upon whether or not he/she adopts new products?” Or, does a person's age bear any relation to brand loyalty behaviour?”. Research questions are too broad to be directly testable. Instead, we reduce the question to one or more hypotheses implied by these questions.
A hypothesis is a conjectural statement of the relation between two or more variables. There are two key characteristics which all hypotheses must have; they must be statements of the relationship between variables and they must carry clear implications for testing the stated relations2. These characteristics imply that it is relationships, rather than variables, being tested. The hypotheses specify how the variables are related and that these must be measurable. Statements lacking any or all of these characteristics are not research hypotheses. Consider the following hypothesis:
“Red meat consumption increases as real disposable incomes increase.”
There is a stated relation between one variable, “red meat consumption”, and another variable, “disposable incomes.” Moreover, both variables are potentially measurable. The criteria are satisfied however for the purposes of statistical testing it is more usual to find hypotheses stated in the so-called null form. The following is an example of a null hypothesis:
‘There is no relationship between red meat consumption and the level of disposable incomes.’
Consider a second hypothesis:
‘There is no relationship between a farmer's educational level and his degree of innovativeness with respect to new farming technologies.’
Again there is a clear statement of the relationship being investigated but there are question marks over the measurability with respect to at least one of the variable i.e. ‘..a farmer's degree of innovativeness’. Researchers may also encounter difficulties in agreeing an appropriate measure of the other variable, i.e. “level of education”. If these problems can be resolved then it may qualify as a hypotheses.
Hypotheses are central to progress in research. They direct the researcher's efforts by forcing him/her to concentrate on gathering the data which will enable the hypotheses to be tested. It is all too easy when conducting research to collect “interesting data” as opposed to “important data”. A second advantage of stating hypotheses is that implicit notions or explanations for events become explicit and this often leads to modifications of these explanations, even before data is collected.
On occasion a give hypotheses may be too broad to be tested. However, if it is a “good” hypothesis then other testable hypotheses may be deduced from it. A problem really cannot be solved unless it is reduced to hypotheses form, because a problem is a question, usually of a broad nature, and is not directly testable.
Marketing research can be carried out on one of three levels; exploratory, descriptive or causal. The researcher needs to advise the person who is commissioning the study on which is the type of research most appropriate to the problem at hand.
Figure 11.3 Three types of research
Exploratory research has the objective of giving a better understanding of the research problem. This includes helping to identify the variables which should be measured within the study. When we have little understanding of the topic we find it impossible to formulate hypotheses without some exploratory research. The techniques of exploratory research include reviews of secondary sources of data, informal interviews and focus group interviews.
To illustrate the point, consider the following case. Crop residues such a straw are high in lignin, a wood like substance, and low in nutrients. This makes it a poor animal feed since the lignin acts against digestibility and the low nutrient content means poor food value. However, if treated in a strong alkali, with the action of internal heat, the lignin breaks down and the nutrient content increases. A company was established to exploit this technology and did so successfully for 4 seasons. After this period sales began to slow down. Three other manufacturers had entered the market by this time. The company, Animal Feed Systems, did not know whether the whole industry had slowed down or if only their product was suffering. Nor did they know if the problem was temporary in that perhaps the market comprised of ‘early adopters’ had been saturated but it was only a matter of time before other farmers began to buy their systems when they saw how well they worked. It was also possible that if a problem did exist it could lie in any one of a number of areas; animal populations might be declining, distributors may not be promoting the product aggressively, customers may be experiencing difficulties in getting the chemicals, and so on and so on.
This is a good example of where insufficient is known to develop clear objectives since the problem cannot be articulated with any precision. Thus the any research would be of an exploratory nature. Such research can take the form of literature searches, informal personal interviews with distributors and users/non-users of the product and/or focus group interviews with prospective customers and/or distributors. Exploratory research is intend to help in the task of formulating a researchable problem and testable hypotheses.
As the name suggests, descriptive research is concerned with describing market characteristics and/or marketing mix characteristics. Typically, a descriptive study specifies the number and size of market segments, the alternative ways in which products are currently distributed, listing and comparison of the attributes and features of competitive products etc.
This type of study can involve the description of the extent of association between variables. For example, it may be observed that there is an association between the geographical location of consumers and their tendency to consume red meat. Note that we are able to describe the relationship rather than explain it. Nonetheless if the relationship between the two is fairly stable this descriptive information may be sufficient for the purposes of prediction. We may, for example, be able to predict how fast the per capita consumption of red meat is likely to rise over a given time period.
The principal difference between exploratory and descriptive research is that, in the case of the latter, specific research question have been formulated before the research is undertaken. When descriptive research is conducted, a great deal is already known about the research problem -perhaps because of a prior exploratory study- and researchers are in a position to clearly define what they want to measure and how to do it.
Causal research attempts to deal with the ‘why’ questions. This type of research is employed when there the objective is to understand to know why a change in one variable brings about a change in another variable. If we can understand the causes of the effects we observe then the ability to predict and control such events is increased.
By way of an illustration consider a common task given to marketing researchers, i.e. that of forecasting sales. Probably the simplest approach is that set of techniques known as time series forecasting. To build a time series forecasting model the researcher will examine historical sales patterns. If the researcher is fortunate there may be a recognisable and recurrent pattern of peaks and troughs in sales when past sales and time periods are set alongside one another. The relationship between the two may be expressed as a mathematical formula or, more simply, could be depicted graphically. The resulting model can be used to forecast sales. No matter how accurate this model proves to be it could not be classified as causal. It would not reveal anything about the reasons ‘why’ sales rises or fall over a given time span. When researchers plot sales against time, as with time series forecasting models, ‘time’ is being used as a proxy variable for unknown explanatory variables. Causal research would seek to identify the individual variables that act, either independently, or in concert to bring about a given effect. A causal forecasting model would incorporate all of these variables and would represent their interactions in the form of a mathematical algorithm.
In addition to deciding who should supply his marketing research needs the manager will also want to contribute to the decision as to what type of data is most appropriate, i.e. primary data or secondary data
The term 'secondary data relates to data which has been collected by individuals or agencies for purposes other than those of a given research study. For example, if a government department has conducted a survey of family food expenditures, then a food manufacturer might use this data in evaluating the total potential market for a new product. Similarly, statistics prepared by a ministry on agricultural production will prove useful to a whole host of people and organisations including, those marketing agricultural supplies, storage companies, transport operators, processing enterprises, commodity brokers, retailers, government policy makers and many more.
No marketing research study should be undertaken without a prior search of secondary sources of data and information. There are several grounds which give confidence to such a bold statement:
Secondary data may be sufficient to solve the problem. On occasion it happens that adequate data may be available to the extent that primary data collection unnecessary.
Data collection costs are substantially lower for secondary data in comparison to primary data. A thorough search of secondary sources can be completed at a fraction of the cost incurred in even a modest primary data collection exercise.
The time involved in searching secondary sources is far less than that needed to complete primary data collection. A systematic search of secondary sources can be completed in a fraction of the time it takes to complete primary data collection.
Secondary sources of information can yield more accurate data than that obtained through primary research. This is not always true but where a government or international agency has undertaken a large scale survey, or even a census, this is likely to yield far more accurate results than independent surveys when these are based on relatively small sample sizes.
Secondary data helps define the research problem and to formulate hypotheses. The assembly and analysis of secondary data almost invariably improves the understanding of the marketing problem, the various lines of inquiry the study could take and the alternative course of action which might be pursued.
Secondary sources help define the population. Secondary data can be extremely useful both in defining the population and in structuring the sample to be taken. For instance, government statistics on a countries' agriculture will help decide how to stratify a sample and, once sample estimates have been calculated, these can be used to project those estimates from the sample to the population.
Whilst the benefits of secondary sources are considerable, their shortcomings have to be acknowledged. The main problems may be categorised as follows:
Definitions: The researcher has to be careful, when making use of secondary data, with regard to the definitions used by those responsible for its preparation. Suppose, for example, the issue of interest is the average family size in rural communities. If published statistics are consulted then a check must be made on how terms such as, family size, have been defined. They may refer only to the nucleus family or could include the extended family. Even apparently simple terms such as ‘farm size’ need careful handling. Such figures may refer to anyone of the following; the land an individual owns, the land an individual owns plus any additional land he rents, the land an individual owns minus any land he rents out, all of his land or only that part of it which he actually cultivates.
Measurement error: Whenever samples are used to estimate population values (e.g. the frequency of purchase of all users based on a sample of users) there are always errors within the estimate. The extent of the error in such estimates is revealed by two statistics, the standard deviation and the standard error of the sampling means. The standard deviation and standard error, these are sometimes not published in secondary sources. The only solution is to try to speak to the individuals involved in the collection of the data to obtain some guidance on the level of accuracy of the data.
Source bias: Researchers have to be aware of vested interests when they consult secondary sources. Those responsible for their compilation may have reasons for wishing to present a more optimistic or pessimistic set of results for their organisation. It is not unknown, for example, food shortages to be exaggerated when reports are being prepared for submission to aid organisations. Similarly, agribusinesses may report lower trading volumes because their levels of trade have implications with respect to tax liability.
Reliability: The reliability of published statistics may vary over time. It is not uncommon, for example, for the systems of collecting data to have changed over time but without any indication of this to the reader of published statistics. Geographical or administrative boundaries may be changed by government or the basis for stratifying a sample may have altered.
Changes in either the method of data collection or in the way variable have been defined also create difficulties when the researcher wishes to describe trends over time or wishes to compare to different time periods.
Time scale: Most census take place at 10 year intervals so data from this and other published sources may be out-of-date at the time the researcher wants to make use of the statistics.
Sources of information
Secondary sources of information may be divided into two categories; internal sources and external sources.
Internal sources of information
All organisations collect information in the course of their everyday operations. Orders are received and delivered, costs are recorded, sales personnel submit visit reports, invoices are sent out, returned goods are recorded and so. Much of this information is of potential use in marketing research but a surprising amount of it is actually used. Organisations frequently overlook this valuable resource by not beginning their search of secondary sources with an internal audit.
For example, consider how much information can be obtained from sales invoices:
sales by territory
sales by customer type
average size of order by customer
customer type, geographical area
average sales by sales person
sales by pack size and pack type.
External sources of secondary information
The main external sources of secondary data are (1) government departments (2) trade associations (3) domestic and international commercial information services (4) national and international development organisations institutions.
Government Statistics: Federal, state and local government departments usually publish a wide range of statistics. These may include all or some of the following:-
social surveys, family expenditure surveys
Trade Associations: Trade Associations differ widely in the extent of their data collection and information dissemination activities. However, it is worth checking with them to determine what they do publish. At the very least one would normally expect that they would produce a trade directory and, perhaps, a yearbook. Chambers of commerce could also prove useful as an information source.
Commercial Services: Published market research report and other publications are available from a wide range of organisations who charge for their information. Typically, marketing people are interested in media statistics and consumer information which has been obtained from large scale consumer or farmer panels. The commercial organisation funds the collection of the data, which is wide ranging in its content, and hopes to profit from selling this data to interested parties.
National and International Institutions: Bank economic reviews, University research reports, journals and articles are all useful sources to contact. International agencies such as World Bank, FAO, UNDP, ITC and ILO produce a plethora of secondary data which can prove extremely useful to the marketing researcher.
Primary research is that which has been specifically designed to address particular marketing problems or questions. Perhaps the approach most readily associated with marketing research is the survey but as we are about to see this is but one, and not invariably the most appropriate, approach. The principal approaches to primary marketing research are:
Survey research: Surveys are characterised by a relatively large number of respondents and the desire to project the results obtained from a sample to a population. If the sample is drawn using a probabilistic method then we can place confidence levels on the inferences we make about the population. Where a non-probabilistic method is used we cannot say how certain or uncertain we are about our inferences. Nonetheless if the sample is reasonably large, and comprised of a good cross-section of the target population, then marketing researchers tend to assume that the sample results are representative of the population. The main forms which surveys take are depicted in figure 11.4.
Figure 11.4 Types of survey
When studies are carried out on a large scale the questionnaires tend to be highly structured. Most, if not all, of the questions will have a closed-end response format. Thus, whilst the large scale survey is the most appropriate approach where the need if for numerical data. For example, if we were considering launching a new range of flavoured milk and wanted to estimate demand in order that we could decide on production schedules then a large scale consumer survey might be the best approach. If on the other hand, our problem were one of finding out why so many people did not drink milk then it would be difficult to design a questionnaire, of reasonable length, which anticipated all possible responses. In such circumstances, it would probably be better to conduct a smaller number of in-depth interviews.
In essence, large scale surveys are useful where the questions are of the 'how many[, ‘how often’ and ‘when’ type but they are blunt instruments for answering questions of the ‘why’ kind.
Qualitative research: In situations where the researcher is primarily interested in why people thing and/or behave in a particular way rather than in being able to quantify things, then qualitative research methods are likely to be employed. Qualitative methods have at least four distinguishing characteristics:
Small numbers of respondents. The idea is to devote a considerable amount of time on each interview to get to the heart of a matter.
Unstructured question formats. That is, the questions are not completely predetermined and the interviewer is free to probe for all details and underlying feelings.
Indirect measurement of respondents' feelings and beliefs. Respondents provide descriptive information about their thought and feelings. These are not easily projected to the population.
Direct observation. The interviewer not only records answers but observes how questions affect interviewees. Hesitant answers, agitation, smiling, sweating, calmness, boredom etc. are all observable and all tell us something about the individuals state of mind.
Three commonly employed qualitative marketing research methods are projective techniques, focus groups and depth interviews.
Figure 11.5 Qualitative research methods
Focus groups: Each focus group generally involves six to eight people who meet with a moderator for a discussion. The discussion is focused, by the moderator on a particular topic. Typically, a group session will last one to two hours. The discussion is free ranging with the moderator intervening only periodically to stimulate the discussion in a particular direction. The moderator uses a discussion guide rather than a questionnaire. This guide is simply an agenda of the topics which the group should cover. Thus, the focus of the discussion, at any point in time, is subtly controlled by the researcher (hence the term moderator). Participants in the groups are chosen on the basis that they belong to the target market.
Any number of focus groups may be held in connection with a particular marketing problem but the results are not strictly projectable to the population since the selection of participants is in no way probabilistic.
Depth interviews: Depth interviews are like lengthy psychoanalytic sessions between a single respondent and a highly skilled interviewer. The idea is to get to the deep, hidden underlying attitudes and feelings the respondent has towards a product, service, company or problems which a product is trying to solve.
Depth interviews are of most value where a study deals with (1) a confidential, emotionally charged or embarrassing matter; (2) a behaviour for which socially acceptable norms exist and the need to conform in group discussions influences responses; (3) a complex behavioural or decision-making process that requires a detailed idiosyncratic, step-by-step description; and (4) when group interviews are difficult to schedule for the target population.
Projective techniques: On occasion, the interests of the research are best served by obtaining information on respondents' beliefs and feelings indirectly. Projective techniques presume that respondents cannot or will not communicate their feelings and beliefs directly. Instead, respondents are encouraged to respond indirectly by projecting their own feelings and beliefs into the situation as they interpret the behaviour of others. The most common projective techniques are:-
|Thematic apperception tests||Respondents are presented with a series of pictures or cartoons in which consumers and products are featured. Participants are asked to study the situation depicted and to comment on what is happening or what might happen next. In this way, respondents are encouraged to project their own feelings and beliefs onto the situation portrayed in the pictures or cartoons. The term thematic apperception test is used because themes (thematic) are elicited based on the perceptual-interpretive (apperception) use of pictures and cartoons.|
|Word association||Respondents are presented with a series of words, one at a time, and asked to indicate what word comes immediately to mind. The respondent's response and time to respond are recorded. Elapsed time and associations are the key measures. Word association is commonly used in the testing of brand names.|
|Sentence completion||Sentence completion tests are similar to word association. Respondents are asked to conclude a number of incomplete sentences with the first word or phrase that comes to mind. Responses are then analysed for content and meaning.|
|Scenario/story completion||Respondents are asked to complete the end of a story or supply the motive for why one or more actors in a story behaved as they did.|
|Third person/role playing||Respondents are presented with a visual or verbal situation in which they are asked to relate the feelings and beliefs of a third person - for example, a friend, neighbour, another farmer or ‘typical’ person - to the situation, rather than to directly express their own feeling/beliefs about the situation. In this way the individual reveals his/her own inner most feelings, attitudes and motives.|
As was said earlier, qualitative research methods are, best employed where the task is to address ‘why’ questions. However, the results of qualitative research are rarely projectable to the population at large. Moreover, they must be carried out by interviewers trained in psychology and/or sociology.
Observation: Methods of data collection involving directly or indirect, human or mechanical measurement of behaviour, are termed observational methods. These can be particularly useful in situations where the respondent is either unable or unwilling to report past behaviour, or in cross-cultural research where it is possible that imperfect translation of the questions can occur. Observation methods are also called naturalistic inquiries because, in their purest form, such studies demand a natural setting. This is because behaviour takes its meaning as much from their context as they do from themselves4. Examples of observation methods include pantry and dustbin audits, and physiological measurements.
Pantry and dustbin audits
In the 1900s an American named Charles Parlin proved to the Campbell Soup Company that wealthy people who employed household help did not use canned soup but that blue-collared families did. His method of establishing his case was to inspect the refuse, or garbage, of a sample of households over a period of time. He reasoned that if simply questioned some respondents would distort their answers because they felt, in those days, that using convenience foods carried a social stigma. Parlin's methods have been extended and refined and today, dustbin audits and pantry audits are commonly used in the industrialised countries. Participating families are asked to place all product packaging in specially marked plastic bags. These are picked up by the research company perhaps twice per week and taken to a central location for checking. Researchers then record, for each item, (1) the product type and class, (2) the number of items of that product, (3) product size (weight for solids, fluid ounces for liquids etc.) (4) price of the product on the container and (5) the brand name of the product. Equally popular as the dustbin audit, and conducted along the same lines, is the audit of participating families' pantry's, fringes and freezers.
A more sophisticated approach to naturalistic inquiries is the monitoring of a respondent's involuntary responses to stimuli. Two types of physiological measurement techniques are the pupilometer and the galvanometer. The pupilometer attaches to a person's head and measures interest and attention by the amount of dilation in the pupil of the eye. It has been used most extensively in the testing of advertisements and product packaging. The galvanometer measures excitement levels by recording the electrical activity in the person's skin.
Figure 11.6 Observation methods
Observational methods are often the most accurate way of measuring overt behaviour but, of course, they have little to contribute in the measurement of attitudes and feelings and other variables which cannot be directly observed.
Experimentation: An experiment entails some sort of test which allows us to discern the effects of an independent variable on a dependent variable. An independent variable is manipulated by the researcher and is sometimes termed and explanatory variable since it is assumed to be related in some way to the dependent variable(s) under study. Experimental approaches to marketing research can be classified as either laboratory experiments or field experiments.
Figure 11.7 Experimental methods
Field experiments: A field experimental environment is a natural setting. For example, suppose the research was intended to measure the effectiveness of alternative merchandising units for a range of dairy products. If there were say two possible designs an experiment could be conducted whereby each design is placed in several retail outlets. Then, over a period of time, a measure of the sales in each outlet could be obtained and a judgement made as to which design, if either, maximised sales. This was the approach used by United Fruits (see case 11.2) when the company was trying to determine what effect the replacement of the Gros Michel variety by the Valery variety would have on retail banana sales. Over a period of time the variety on retail shelves was switched back and forth and sales levels were recorded. The reader may recall that the conclusion from this test was that the variety had very little effect on sales; suggesting consumers could not tell the difference between the two bananas.
Control is an important factor in all forms of experimentation, including field experiments. In the case of the example of the dairy products merchandising units great care would have to be taken to match the retail outlets each of the merchandising units is placed in on criteria like their size, location etc. In short, we wish to be sure that any differences or variance in sales can be confidently put down to the merchandising unit and not to differences in other aspects of the retail outlets.
Field experiments can be applied to may marketing problems but perhaps the most noted application is that of test markets. Prior to a company or organisation undertaking a national launch of a new product/service it may be decided to launch not only the product but the proposed marketing programme in a limited area. The area is normally selected on the basis of its being a microcosm of the country. Test markets, it should be remembered, are used to test the whole of the proposed marketing mix. If the results of the test market warrant an adjustment to elements of the marketing mix then this can be done before the national launch.
Laboratory experiments: The laboratory environment allows the researcher to have direct control over most, if not all, of the crucial factors that might affect the experimental outcome. Laboratory experiments have been used in concept testing, taste testing, packaging testing, advertising effectiveness studies and simulated test markets.
The chief advantage of experimental approaches is that behaviour can be directly observed rather than asking people about events retrospectively, when their memory may not be accurate; or ask them to project how they would behave in a given set of circumstances, when it is difficult for them to be sure of how they would react to an actual phenomenon or stimulus. Experiments however, do not give rise to statistical generalisation. Whereas a researcher might ask 1,000 people, or more, what they think of a given product, and then project these results to the population, few would consider running the same experiment 1,000 times or more. Experiments give rise to analytic generalisation to a theory and not statistical generalisation to a population.
Continuous research: Certain types of data are gathered on a regular basis as opposed to the ad hoc survey. Moreover, researchers will use standardised methods in order that the data collected at one point in time is comparable with that collected at other times. In this way, a picture of market trends can be built up. This type of longitudinal research is often funded on a syndicated basis. Syndicated research usually involves an independent research company collecting data and supplying it simultaneously to a number of clients. For example, the same group of farmers may be sent a standardised questionnaire every 6 months with a battery of questions about what implements they have purchased since the last survey, what herbicides, pesticides, animal health products and or post harvest treatments they have purchased, what application rates they have used with respect to seeds, crop/animal protection chemicals etc. This data would then be sold to a wide range of companies supplying agricultural inputs. These companies are likely to be regular, rather than ad hoc, subscribers to the research. Reports to these subscribers will be customised to some degree. For example, reports can be organised on the basis of the client's sales territories and, of course, a seed company would not, for example, be given a report containing data on animal health products. Figure 11.9 depicts the principal types of continuous research studies.
Figure 11.8 Types of continuous research
Diary panels involve samples of households that have agreed to provide specific information regularly over an extended period of time. For this reason they are often referred to as continuous panels. Respondents are asked to keep a specially designed dairy. Purchase panels record details of the products they purchased, sizes, brands, flavours, prices paid etc. Media panels record details of newspapers, magazines, periodicals bought and/or read, television programme watched, radio stations listened to and so on. Typically, the completed diary is returned to the research company every one to four weeks.
Media panels are primarily used for establishing advertising rates for radio, television and printed media. Purchase panel data can be used to forecast sales levels or market shares of new products, for identifying trends and establishing demographic profiles of specific user groups, for evaluating test markets, for testing different advertising campaigns and for estimating brand switching and repeat purchase rates.
Although the discussion has focused on consumer panels it should be noted that panels can, and have been, successfully established using farmers.
Audit services: An audit involves a systematic examination of either how much of a product has sold at the shop/store level (retail audit) or how much of a product has been withdrawn from warehouses and delivered to retailers (warehouse withdrawal audits). Participating wholesale/retail outlets receive basic reports and a cash payment. Like panel data, the figures compiled from these exercises are sold to a large number of clients; many of whom compete against one another.
Audits provide relatively precise information on the movement of many different types of goods. Since most products are not sold directly to the end user, but to retailers, wholesalers and distributors, the manufacturer does not have information on sales at the retail level. Even though information on factory shipments are available, warehouse stocks might be accumulating because of limited retail sales. Moreover, audits give information on how competing products are faring in the marketplace.
Continuous research provides a type of information not available through and hoc surveys, experiments or observation exercises i.e. trends in consumption, tastes and fashion. If a manufacturer is a regular subscriber to this type of data he/she can see patterns in the marketplace emerge and take pre-emptive action. However, there are potential weaknesses in continuous research methods. With panels the main problems are threefold; (1) they often under-represent minority groups in the population, (2) over time the panel ages and so there are heavy costs in continually recruiting new, younger participants to maintain the panel's representativeness, and (3) knowing their purchases or readership/viewing/listening behaviour is being scrutinized can alter that behaviour. In the case of audits the chief hazard is that if a large wholesale/retail organisation, or outlet, refuses to participate then the sample may be far from reliable for the purposes of projection. Another problem with audits is that they are time consuming. There is typically a two month gap between data collection and the publication of the report
Those new to marketing research often intuitively believe that decisions about the techniques of analysis to be used can be left until after the data has been collected. Such an approach is ill-advised. Before interviews are conducted marketing researchers should be apply the following check-list:
Do I know how each and every question is to be analysed? (e.g. which univariate or bivariate descriptive statistics, tests of association, parametric or nonparametric hypotheses tests, or multivariate methods are to be used?)
Do I have a sufficiently sound grasp of these techniques to apply them with confidence and to explain them to the decision-maker who commissioned the study?
Do I have the means to perform these calculations? (e.g. access to a computer which has an analysis program with which I am familiar? Or, if the calculations have to be performed manually, is there sufficient time to complete them and then to check them?)
If a computer program is to be used at the data analysis stage have the questions been properly coded?
Have I scaled the questions correctly for the chosen statistical technique? (e.g. A t-test cannot be used on data which is only ordinal or ranked)
There is little point in spending time and money on collecting data which subsequently is not or cannot be analysed. Therefore consideration has to be given to issues such as these before the fieldwork is undertaken.
At this stage the researchers are ready to go into the field and collect the data. Before deploying interviewers in the field there are several aspects of recruitment and fieldwork which have to be carefully managed. First, the manager must be aware that interviewers are required to achieve, and maintain, a high standard of work without continuous supervision or daily contact with colleagues. The manager must also be conscious of the fact that the task of the interviewer is not mechanistic: interviewers often have to use a great deal of initiative and effort, for example in locating obscure addresses, securing appointments and co-operation from reluctant -perhaps suspicious- respondents and administer complex questionnaires. Third, the manager may have to take into account that, although most professional survey organisations have full-time teams of trained interviewers, additional, often untrained, interviewers frequently have to be recruited to supplement the trained field force. For ad hoc surveys, in some instances, a marketing manager may decide to train, organise and supervise the survey team him/herself.
All of this underlines the fact that fieldworkers have to be fully trained before they go into the field. Moreover, a manager has to be selective when employing field staff since their task is technically skilled and requires a high level of dedication and a tenacious spirit if the work is to be properly completed.
Above all, data collection has to be well planned. In addition to ensuring that field personnel are adequate to the task both in number and training, care has to be taken in the planning of sampling, call backs on respondents absent on the first visit, logistics with respect to data gatherers and the collection and submission of questionnaires, checking of completed data forms/questionnaires, data analysis and so on.
The word ‘analysis’ has two component parts, the prefix ‘ana’ meaning ‘above’ and the Greek root ‘lysis’ meaning ‘to break up or dissolve’. Thus, Dey6 describes data analysis as:
“…a process of resolving data into its constituent components, to reveal its characteristic elements and structure.”
Where the data is quantitative there are three determinants of the appropriate statistical tools for the purposes of analysis. These are the number of samples to be compared, whether the samples being compared are independent of one another and the level of data measurement.
Suppose a fruit juice processor wishes to test the acceptability of a new drink based on a novel combination of tropical fruit juices. There are several alternative research designs which might be employed and each involving different numbers of samples.
|Test A||Comparing sales in a test market and the market share of the product it is targeted to replace.||Number of samples = 1|
|Test B||Comparing the responses of a sample of regular drinkers of fruit juices to those of a sample of non- fruit juice drinkers to a trial formulation,||Number of samples = 2|
|Test C||Comparing the responses of samples of heavy, moderate and infrequent fruit juice drinkers to a trial formulation||Number of samples = 3|
The next consideration is whether the samples being compared are dependent (i.e. related) or independent of one another (i.e. unrelated). Samples are said to be dependent, or related, when the measurement taken from one sample in no way affects the measurement taken from another sample. Take for example the outline of test B above. The measurement of the responses of fruit juice drinkers to the trial formulation in no way affects or influences the responses of the sample of non-fruit juice drinkers. Therefore, the samples are independent of one another. Suppose however a sample were given two formulations of fruit juice to taste. That is, the same individuals are asked first to taste formulation X and then to taste formulation Y. The researcher would have two sets of sample results, i.e. responses to product X and responses to product Y. In this case, the samples would be considered dependent or related to one another. This is because the individual will make a comparison of the two products and his/her response to one formulation is likely to affect his/her reaction or evaluation of the other product.
The third factor to be considered is the levels of measurement of the data being used. Data can be levels of measurement:nominal, levels of measurement:ordinal, levels of measurement:interval or levels of measurement:ratio scaled. The following table summarises the mathematical properties of each of these levels of measurement.
Table 11.1 Levels of measurement
|Measurement scale||Measurement Level||Examples||Mathematical properties|
|Nominal||Frequency counts||Produce grading categories||Confined to a small number of tests using the mode and frequency.|
|Ordinal||Ranking of items||Placing brands of cooking oil in order of preference||Wide range of nonparametric tests which test for order.|
|Interval||Relative differences of magnitude between items||Scoring products on a 10 point scale of like/dislike||Wide range of parametric tests|
|Ratio||Absolute differences of magnitude||Stating how much better one product is than another in absolute terms.||All arithmetic operations|
Once the marketing researcher knows how many samples are to be compared, whether these samples are related or unrelated to one another and the level of measurement then the selection of the appropriate statistical test is easily made. Figure 11.9 provides a useful guide to making that final selection.
To illustrate the importance of understanding these connections consider the following simple, but common, question in marketing research. In many instances the age of respondents will be of interest. This question might be asked in either of the two following ways:
|Please indicate to which of the following age categories you belong-|
|15–21 years||How old are you? Years|
|Over 30 years|
Choosing format (a) would give rise to nominal (or categorical) data and format (b) would yield ratio scaled data. By These are at opposite ends of the hierarchy of levels of measurement. If by accident or design format (a) were chosen then the analyst would have only a very small set of statistical tests that could be applied and these are not very powerful in the sense that they are limited to showing association between variables and could not be used to establish cause-and-effect. Format (b), on the other hand, since it gives the analyst ratio data, allows all statistical tests to be used including the more powerful parametric tests whereby cause-and-effect can be established, where it exists. Thus a simple change in the wording of a question can have a have a fundamental effect upon the nature of the data generated.
Figure 11.9 Selecting statistical tests
The individual responsible for commissioning the research may be unfamiliar with the technicalities of statistical tests but he/she should at least be aware that the number of samples, their dependence or independence and the levels of measurement does affect how the data can be analysed. Those who submit marketing research proposals, involving quantitative data, should demonstrate an awareness of the factors that determine the mode of analysis and a capability to undertake such analysis.
The end products of marketing research are conclusions and recommendations. Marketing research should be designed to help to identify potential threats and opportunities, generate alternative courses of action, provide information to enable marketing managers to evaluate those alternatives and advises on the implementation of the alternatives.
Moser and Kalton7 believe that:
“…whatever the nature of the data, the task of interpretation falls squarely on the shoulders of the researcher.”
Not everyone agrees with this perspective. Some believe that researchers should confine themselves to ‘reporting the facts’. Ehrenberg8 says that this is not only undesirable but is impossible since implicity or explicitly, the researcher's value judgements will colour the presentation of ‘the facts’. The view taken in this textbook is that researchers should interpret their data. At worst, the researcher will at least display his/her biases and prejudices and at best he/she will share valuable insights gained as a direct result of carrying out the fieldwork. Too often marketing research reports chiefly comprise a lengthy series of tables of statistics accompanied by a few brief comments which verbally describe what may already be self-evident from the tables. Without interpretation, data remains data rather than information. It is information which management needs to reduce the inherent risks and uncertainties in management decision making.
Guidelines on report content and presentation
The results of marketing research must be effectively communicated to management and the commissioner of the research should, on behalf of him/herself, and the marketing colleagues who intend to act on the report, that is both clear and concise. Presenting the results of a marketing research study to management generally involves a formal written report as well as an oral presentation. The report and presentation are extremely important. First because the results of marketing research are often intangible (after the study has been completed and a decision is made there is very little physical evidence of the resources such as time and effort, that went into the project); the written report is usually the only documentation of the project. Second, the written report and the oral presentation are often the only aspect of the study to which marketing executives are exposed, and consequently the overall evaluation of the research project rests on how well this information is communicated. Third, since the written research report and oral presentation are typically the responsibility of the marketing research supplier, the effectiveness with which the results are communicated and the usefulness of the information provided plays a crucial role in establishing the reputation of the researcher.
Preparing a research report involves other activities besides writing; in fact, writing is actually the last step in the preparation process. Before writing can take place, the results of the research project must be fully understood and thought must be given to what the report will say. Thus, preparing a research report involves three steps: understanding, organising and writing. The general guidelines that researchers should be encouraged to followed for any report are as follows:
|1) Think of the audience||The information resulting from the study is ultimately of importance to marketing managers, who will use the results to make decisions Thus, the report has to be understood by them; the report should not be too technical and jargon avoided wherever possible. For example, where statistical tests have been applied it should not be assumed that the reader understands terms such as significance level, degrees of freedom, type 1 errors etc. Rather, where necessary these should be explained in non-technical terms.|
|2) Be concise yet complete||On the one hand, a written report should be complete in the sense that it stands by itself and that no additional clarification is needed. On the other hand, the report must be concise and must focus on the critical elements of the project and must exclude issues that are not material to the decisions that the managers wishes to be in a position to make.|
|3) Understand the results and draw conclusions:||The managers who read the report are expecting to see interpretive conclusions in the report. The researcher must therefore understand the results and be able to interpret these. Simply reiterating facts won't do, and the researcher must ask him/herself all the time “so what”; i.e. so what are the implications of this particular result.|
The summary of findings is perhaps the most important component of the written report, since many of the management team who are to receive a copy of the report will only read this section. The summary of findings is usually put right after the title page, or is bound seperately and presented together with the report.
The introduction should describe the background of the study and the details of the research problem. Following that, automatically the broad aim of the research can be specified, which is then translated into a number of specific objectives. Furthermore, the hypotheses that are to be tested in the research are stated in this section.
In the methodology chapter the sampling methods and procedures are described, as well as the different statistical methods that are used for data analysis. Finally, the sample is described, giving the overall statistics, usually consisting of frequency counts for the various sample characteristics.
The figure below contains a suggested outline format for writing the research report.
Figure 11.10 Research report writing format
Once the sample has been described, the main findings are to be presented in such a way that all objectives of the study are achieved and the hypotheses are tested. As mentioned before, it is essential that the main findings are well interpreted and conclusions are drawn wherever possible.
Wherever possible the research proposal should contain a skeleton outline of the contents of the final marketing research report. That is, the researcher should convey to the decision maker the kinds of information that he/she intends to put into the report. If this is not exactly what the decision maker feels he/she requires then the differences can be resolved at an early stage of the project rather than becoming a source of conflict after the research has been completed.
Marketing research helps reduce the level of uncertainty with which marketing managers must cope by converting the raw facts and figures of data into information. In order to do this effectively, marketing research must be conducted systematically, objectively and analytically.
Decision makers wishing to make use of marketing research must communicate the purpose and objectives of the research, the resources which can be devoted to a particular study and the time constraints of the study.
The individual or group charged with undertaking a marketing research exercise responds to the research brief with a research proposal. The research proposal sets out the research design and the methodology that the researcher proposes to follow. There are eight essential steps in research design; problem defination, determination of whether the study required is exploratory, descriptive or causal, selection of the method(s) of data gathering, a plan of the data will be analysed, an outline plan of the data collection programme and a skeleton outline of what the final report will contain.
Marketing research findings and conclusions must be effectively communicated to management. Written reports need to be both clear and concise. A sensitivity towards the needs of the reader is of prime importance. The report should not be too technical and jargon avoided wherever possible, must exclude issues that are not material to the decisions that the managers wishes to be in a position to make and should be rich in interpretive conclusions.
|Audits||Exploratory research||Qualitative research|
|Causal research||Hypotheses||Research brief|
|Continuous research||Measurement levels||Research design|
|Descriptive research||Observation methods||Secondary research|
|Experimentation||Primary research||Syndicated research|
From your knowledge of the material in this chapter, give brief answers to the following questions below.
Name the 3 key words used in the definition of marketing research by Green, Tull and Albaum.
Define the term hypothesis.
What are the 3 types of research described in this chapter?
Under what circumstances would depth interviews be an appropriate method of data collection?
What are the principal problem of operating research panels?
What factors determine which statistical tests are appropriate for a given data set?
What are the key characteristics of qualitative research?
What are the 2 forms of experimentation?
List the main problems to be aware of when making use of secondary data.
Explain how the thematic apperception test method is applied in marketing research.
What is the advantage of drawing probabilistic samples?
What term is used to indicate the leader of a focus group session?
1. Green, P.E., Tull, D.S. and Albaum, G. (1988), Research For Marketing Decisions, 5th edition, Prentice-Hall, New Jersey, P.2.
2. Kerlinger, F.N. Foundations Of Behavioral Research, (19**) 2nd ed. Holt, Rhinehart and Winston, NewYork, P. 18.
3. Fox, J.M. (1985), “How ‘Chiquita’ Helped United Fruit”, In: Successful Agribusiness, J. Freivalds (Ed.) Gower Publishing Company, Aldershot, pp.112–129.
4. Dillon, W.R., Madden, T.J. and Firtle, N.H. (1990), Marketing Research in A Marketing Environment, 2nd edition, Irwin, P. 163–168.
5. Dixie, G. (1989), Horticultural Marketing, FAO Agricultural Service Bulletin 76, Food and Agriculture Organization of the United Nations, Rome, pp.22–24.
6. Dey, I. (1993), Qualitative Data Analysis: A User-Friendly Guide For Social Scientists,
7. Moser, C.A. and Kalton, G. (1971), Survey Methods In Social Investigations, Heinnemann, London.
8. Ehrenberg, A.S.C. (1964), “What Research For What Problem?”, In: Research In Marketing, Market Research Society, London.