The Market Research Process

By Hair, J.F., Bush, R.P., Ortinau, D.J.

Edited by Paul Ducham


Before we introduce and discuss the phases and specific steps in the information research process, it is important that you understand when research is needed and when it is not.

While many marketing research texts suggest the first step in the marketing research process is for the researcher to establish the need for marketing research, this places too much responsibility and control in the hands of a person who might not be trained in understanding the management decision-making process. Decision makers and researchers frequently are trained differently in their approach to identifying and solving business problems, questions, and opportunities, as illustrated in the nearby A Closer Look at Research (In the Field) box. Until decision makers and marketing researchers become closer to each other in their thinking, the initial recognition of a problem or opportunity should be the primary responsibility of the decision maker, not the researcher.

Nevertheless increasingly researchers must interact closely with managers in recognizing and identifying business problems and opportunities. Decision makers often initiate the research process because they recognize that more information is needed before a good plan of action can be developed. It is at this point, especially, that the help and advice of the researcher are required. Once the research process is initiated, in most cases decision makers will need assistance in defining the problem, collecting and analyzing the data, and interpreting the data. But when is it advisable to initiate the research process? For decision makers given the responsibility of recognizing and defining a problem or opportunity, a good rule of thumb is to ask, “Can the decision-making problem (or question) be resolved based on past experience and managerial judgment?” Only if the response is “no” should research be considered and perhaps implemented.

Akey to knowing when the information research process should be undertaken is to understand that marketing research no longer focuses on primary data to solve management’s problems. Increasingly, secondary research and data warehouse information are being used to address decision making. Technological advances in the Internet, high-speed communication systems, and faster secondary and primary data acquisition and retrieval systems are dramatically changing marketing research practices in that more problems are being resolved with secondary data instead of collecting primary data.

Four situations in which the decision to commission a marketing research project may be ill advised are when sufficient secondary information already is available, when time is of the essence, when resources are inadequate, and when costs of research are too high (see Exhibit 2.2). However, an important caveat to these cautions is that the decision maker is assumed to have precise knowledge about the true availability of existing information, the necessary time, staff, and adequate resources should a research project be initiated, and the expected value or cost/benefit trade-off of the resulting information. But such knowledge is a rare thing. And as technology advances, bearing on all the factors in decision making unpredictably, this assumption becomes more not less suspect, complicating the decision still further. For instance, the time and money needed to conduct an interview may go down with new technology, even while the availability of secondary data increases. Here the research expert can be of great assistance to managers trying to decide the nature of the research effort to be conducted or whether or not sufficient internal information exists already.

Exhibit 2.3 displays a framework for determining whether the research process is necessary. To repeat, the initial responsibility of the decision maker is to determine if research should be used to collect the needed primary information. Or can the problem or opportunity be resolved using existing (secondary) information and managerial judgment? The focus is on deciding what type of information (secondary or primary) is really required to answer the research question(s). In reality, conducting either secondary or primary research costs time, effort, and money. So the bottom line is that if there is an opportunity, but decision makers do not have the right information or are unwilling to rely on the information at hand, a research effort may be warranted.

A related question is: Does the problem/opportunity have strategic or significant tactical importance? Strategic decisions generally have longer time horizons and are more complex than tactical decisions. Most strategic decisions are critical to the company’s profit objectives, but tactical decisions can be important as well. For example, Outback Steakhouse recently made a tactical decision to update its menu both in appearance and food offerings. Researching the opinions of customers proved very helpful in determining new food items to be included and items that should be offered as occasional “chef’s specials.” Thus, if the problem has strategic or significant tactical importance, a research expert should be consulted and perhaps a research process implemented.

With the assistance of the research expert, decision makers are in a better position to answer the question: Is adequate information available within the company’s internal record systems to resolve the problem? In the past, if the necessary marketing information was not available in the firm’s internal record system, then a customized marketing research project was undertaken to obtain the information. The resources today available on the Internet may allow a problem to be solved with secondary data.

With input from the research expert, decision makers must assess the “time constraints” associated with the problem/opportunity: Is there enough time to conduct the necessary research before the final managerial decision must be made? Today’s decision makers often need information in real time. But in many cases, systematic research that delivers highquality information can take months. If the decision maker needs the information immediately, there may not be enough time to complete the research process. Another fundamental question focuses on the availability of marketing resources such as money, staff, skills, and facilities. Many small businesses lack the funds necessary to consider doing formal primary research even if it would be valuable.

Acost-benefit assessment should ask: Do the benefits of having the additional information outweigh the costs of gathering the information? While the costs of doing marketing research vary from project to project, generally they can be estimated accurately. But predetermining the benefits and true value of the expected information remains difficult.

Other questions to consider before starting a research project include:

  • What is the perceived importance and complexity of the problem?
  • Is the problem realistically researchable? Can the critical variables in the proposed research be adequately designed and measured?
  • Will conducting the needed research give valuable information to the firm’s competitors?
  • Will the research findings be implemented?
  • Will the research design and data represent reality?
  • Will the research results and findings be used as legal evidence? 
  • Is the proposed research politically motivated?

Finally, it is useful to go back to the firm’s broadest strategic considerations in answering the most basic question: Why should the decision maker conduct information research?

  1. If the information will clarify the problem or identify marketplace changes that directly influence the company’s product/service responsibilities.
  2. If the information helps the company to acquire meaningful competitive advantages within its market environment.
  3. If the information leads to marketing actions that will achieve marketing objectives.
  4. If the information provides proactive understanding of future market conditions.

If the research process will likely throw good light on any of these, then it is probably worth it.



The primary goal of the research process is to provide decision makers with knowledge that will enable them to resolve problems or pursue opportunities. Knowledge is created only after the data have been collected, analyzed, and interpreted so decision makers can make decisions.

To understand this process, decision makers need to know the difference between raw data, data structures, information, and knowledge. First, raw data are the actual responses that are obtained about an object or topic of investigation by asking questions or observing actions. These initial responses have not been analyzed or given an interpretive meaning. Some examples of raw data are (1) the responses on a questionnaire; (2) the words recorded during a focus group interview; (3) the number of vehicles that pass through a specified intersection; and (4) the list of purchases, by product type, recorded by an electronic cash register at a local supermarket.

All secondary and primary marketing information is derived from the following process: gather raw responses; apply some form of data analysis to create usable data structures; and then have someone (a researcher or decision maker) interpret those data structures.

Gather raw data → Create data structures → Provide interpretation

Data structures are the result of combining individual raw responses into groups of data using some type of quantitative or qualitative analysis procedure (e.g., content analysis, calculation of sample statistics). The results reveal data patterns or trends, which in turn can be simple or complex. Some examples are (1) the average number of times 500 moviegoers patronize their favorite movie house; (2) the frequency distribution of 1,000 college students eating at several predetermined restaurants in a 30-day time frame; (3) the sampling error associated with the overall satisfaction of 250 new Acura 3.2TL automobile owners; and (4) the analysis of variance statistical results of comparing hotel selection criteria means for first-time and repeat customers of a particular hotel.

Information becomes knowledge when someone (either the researcher or the decision maker) interprets the data and attaches meaning. To illustrate this process, consider the Excelsior Hotel. Corporate executives were assessing ways to reduce costs and improve profits. The VP of finance suggested cutting back on the “quality of the towels and bedding” in the rooms. Before making a final decision, the president asked the marketing research department to interview business customers.

Exhibit 2.5 summarizes the key results. Atotal of 880 people were asked to indicate the degree of importance they placed on seven criteria when selecting a hotel. Respondents used a six-point importance scale ranging from “Extremely Important = 6” to “Not At All Important = 1.” The average importance of each criterion was calculated for both first-time and repeat customers and statistically significant differences were identified. These results do not confirm, however, whether “quality towels and bedding” should be cut back to reduce operating costs.

When shown the results, the president asked this question: “I see a lot of numbers, but what are they really telling me?” The director of marketing research quickly responded by explaining: “Among our first-time and repeat business customers, the quality of the hotel’s towels and bedding is considered one of the three most important selection criteria impacting their choice of a hotel to stay at when an overnight stay is required. In addition, they feel cleanliness of the room and offering preferred guest card options are of comparable importance to the quality of towels and bedding. But, firsttime patrons place significantly higher importance on cleanliness of the room than do repeat patrons (5.75 vs. 5.50). Moreover, repeat customers place significantly more importance on the availability of our preferred guest card options than do business patrons (x = 5.71 vs. 5.42).”

Based on these comments, the executives decided they should not cut back on the quality of towels or bedding as a way to reduce expenses and improve profitability.



Exhibit 2.6 shows the steps included in each phase of the research process. Although in many instances researchers follow the four phases in order, individual steps may be shifted or omitted. The complexity of the problem, the urgency of solving the problem, the cost of alternative approaches, and the clarification of information needs will directly impact how many of the steps are taken and in what order. For example, secondary data or “off-theshelf ” research studies may be found that could eliminate the need to collect primary data. Similarly, pretesting the questionnaire (step 7) might reveal weaknesses in some of the scales being considered (step 6), resulting in further refinement of the scales or even selection of a new research design (back to step 4).

What might happen if the research process is not followed? Substantial time, energy, and money can be spent with the result being incomplete, biased, or wrong information. For example, the Food and Beverage Committee at the Alto Lakes Golf and Country Club in Alto, New Mexico, wanted to determine members’ satisfaction with the “beverage cart” services being provided on the golf course and learn how to improve services. Not knowing the research process, the committee asked members to rate the beverage cart service using a six-point scale ranging from (6) “Outstanding” to (1) “Terrible” and provided space for written comments. After reviewing only 50 cards returned, the committee found that some members’ comments were related to beverage cart satisfaction. But the rating scale was measuring performance rather than satisfaction. The committee did obtain information about the beverage cart service, but it was not what they wanted. Although the committee included some of the key activities in the research process, the data did not answer their questions.



Generally, decision makers prepare a statement of what they believe is the problem before the researcher becomes involved. Then researchers assist decision makers to make sure the problem or opportunity has been correctly defined and the information requirements are known.

For researchers to understand the problem, they use a problem definition process such as shown in Exhibit 2.7. There is no one best process. But any process undertaken should include the following activities: (1) agree on the decision maker’s purpose for the research, (2) understand the complete problem, (3) identify measurable symptoms, (4) select the unit of analysis, and (5) determine the relevant variables. Correctly defining the problem is an important first step in determining if research is necessary. A poorly defined problem can produce research results that are of little value, as in the New Coke example in the nearby A Closer Look at Research (In the Field) box.

Purpose of the Research Request

Problem definition begins by determining the research purpose. Decision makers must decide whether the services of a researcher are needed. Then, the researcher begins to define the problem by asking the decision maker why the research is needed. Through questioning, researchers begin to learn what the decision maker believes the problem is. Having a general idea of why research is needed focuses attention on the circumstances surrounding the problem. Using the iceberg principle displayed in Exhibit 2.8 helps researchers to distinguish between the symptoms and the causes.

Understand the Complete Problem Situation

The decision maker and the researcher both must understand the complete problem. This is easy to state but quite often difficult to execute. To gain an understanding of the complete problem, researchers and decision makers should do a situation analysis of the problem. A situation analysis is a tool that focuses on gathering background information to familiarize the researcher with the overall complexity of the problem. A situation analysis attempts to identify the events and factors that have led to the situation, as well as any expected future consequences. Awareness of the complete problem situation provides better perspectives on the decision maker’s needs, the complexity of the problem, and the factors involved.

A situation analysis enhances communication between the researcher and the decision maker. The researcher must understand the client’s business, including factors such as the industry, competition, product lines, markets, and in some cases production facilities. To do so, the researcher cannot rely solely on information provided by the client because many decision makers either do not know or will not disclose the information needed. Only when the researcher views the client’s business objectively can the true problem be clarified.

Identify and Separate Out Symptoms

Once the researcher understands the overall problem situation, he or she must work with the decision maker to separate the possible basic problems from the observable and measurable symptoms that may have been initially perceived as being the problem. For example, many times managers view declining sales or loss of market share as problems. After examining these issues, the researcher may see that they are really symptoms—the result of more specific issues such as poor advertising execution, lack of sales force motivation, or inadequate distribution. The challenge facing the researcher is one of clarifying the real problem by separating out possible causes from symptoms. Is a decline in sales truly the problem or merely a symptom of lack of planning, poor location, or ineffective sales management?

Determine the Unit of Analysis

As a fundamental part of problem definition, the researcher must determine the appropriate unit of analysis for the study. The researcher must be able to specify whether data should be collected about individuals, households, organizations, departments, geographical areas, or some combination. The unit of analysis will provide direction in later activities such as scale development and sampling. In an automobile satisfaction study, for example, the researcher must decide whether to collect data from individuals or from a husband and wife representing the household in which the vehicle is driven.

Determine the Relevant Variables

The researcher and decision maker jointly determine the variables that need to be studied. The types of information needed (facts, predictions, relationships) must be identified. Exhibit 2.9 lists examples of variables that are often investigated in marketing. Variables are often measured using several related questions on a survey and may be called constructs. In some situations we refer to these variables as constructs.




Next, the researcher must reformulate the problem in scientific terms. That is, the researcher must redefine the problem as a research question because the scientific approach ensures a systematic approach to developing problem solutions. For the most part, this is the responsibility of the researcher. To provide background information on other firms that may have faced similar problems, the researcher conducts a review of the literature.

Breaking down the problem into research questions is one of the most important steps in the marketing research process because how the research problem is defined influences all of the remaining research steps. The researcher’s task is to restate the initial variables associated with the problem in the form of key questions: how, what, where, when, and why. For example, management of Lowe’s Home Improvement Inc. was concerned about the overall image of Lowe’s retail operations as well as its image among customers within the Atlanta metropolitan market. The initial research question was “Do our marketing strategies need to be modified to increase satisfaction among our current and future customers?” After Lowe’s management met with consultants at Corporate Communications and Marketing, Inc., to clarify the firm’s information needs, the consultants translated the initial problem into the specific questions displayed in Exhibit 2.10. With assistance of management, the consultants then identified the attributes in each research question. For example, specific “store/operation aspects” that can affect satisfaction included convenient operating hours, friendly/courteous staff, and wide assortment of products and services.

When research questions are written, two approaches can be taken to determine the level of detail to use. One approach is to phrase the questions to include only the general category of possible factors. For example, with the demographic question in the Lowe’s example (Exhibit 2.10), the phrasing is somewhat ambiguous because it expresses only the need for a “demographic/psychographic profile” of customers without specifying which particular demographic characteristics (e.g., age, income, education level, marital status) or psychographic factors (e.g., price conscious, do-it-yourself, brand loyalist, information seeker) should be investigated. The other approach is to be more specific in phrasing the research questions. For example, if Lowe’s management team is interested in determining the price range for a particular Black & Decker power drill, the research questions would be phrased as follows: “What are the price ranges customers expect to pay for the Black & Decker RX power drill?” and “What are the price ranges customers are willing to pay for the Black & Decker RX power drill?” Here each research question focuses on a specific data requirement—expected price ranges and then actual price ranges.

After redefining the problem into research questions and identifying the information requirements, the researcher must determine the types of data (secondary or primary) that will best answer each research problem. Although a final decision on types of data is part of Step 4 (Determine the Research Design and Data Sources), the researcher begins the process in Step 2. The researcher asks the question “Can the specific research question be addressed with data that already exist or does the question require new data?” To answer this question, researchers consider other issues such as data availability, data quality, and budget and time constraints.

Finally, in Step 2 the researcher determines whether the information being requested is necessary. This step must be completed before going on to Step 3.



The research objectives should be based on the definition of the research problem in Step 2. Formally stated research objectives provide guidelines for determining other steps that must be taken. The assumption is if the objectives are achieved, the decision maker will have the information needed to solve the problem.

Research objectives serve as the justification for management and researchers to undertake a research project. Consider the Ford Foundation example in the nearby A Closer Look at Research (Small Business Implications) box. Notice that the objectives listed at the end are different from the foundation’s statement of the research problem and the researcher’s problem. Before researchers move beyond Phase I of the research process, they must make sure each factor in the study has been defined. There also must be justification for the relevance of each factor. For example, what does the Ford Foundation really mean by “protection”? Protection from what—rain or cold or snow, or perhaps something else?

Before moving to Phase II of the research process, the decision maker and the researcher must evaluate the expected value of the information. This is not an easy task because a number of factors come into play. “Best guess” answers have to be made to the following types of questions:

“Can the information be collected at all?”

“Can the information tell the decision maker something not already known?”

“Will the information provide significant insights?”

“What benefits will be delivered by this information?”

In most cases, research should be conducted only when the expected value of the information to be obtained exceeds the cost.


The research design serves as an overall plan of the methods used to collect and analyze the data. Determining the most appropriate research design is a function of the research objectives and information requirements. The researcher must consider the types of data, the data collection method (for example, survey, observation, in-depth interview), sampling method, schedule, and budget. There are three broad categories of research designs: exploratory, descriptive, and causal. An individual research project may sometimes require a combination of exploratory, descriptive, and/or causal techniques in order to meet research objectives.

Exploratory research has one of two objectives: (1) generating insights that will help define the problem situation confronting the researcher or (2) deepening the understanding of consumer motivations, attitudes, and behavior that are not easy to access using other research methods. Examples of exploratory research methods include literature reviews of already available information; qualitative approaches such as focus groups and In-Depth Interviews; or pilot studies.

Descriptive research involves collecting numeric data to answer research questions. Descriptive information provides answers to who, what, when, where and how questions. In marketing, examples of descriptive information include consumer attitudes, intentions, preferences, purchase behaviors, evaluations of current marketing mix strategies, and demographics.

Descriptive studies may provide information about competitors, target markets, and environmental factors. For example, many chain restaurants conduct annual studies that describe customers’ perceptions of their restaurant as well as perceptions of primary competitors. These studies, referred to as either image assessment surveys or customer satisfaction surveys, describe how customers rate different restaurants’ customer service, convenience of location, food quality, and atmosphere. Some qualitative research is descriptive in the sense of providing rich or “thick” narrative description of phenomena. However, the term “descriptive research” usually means numeric rather than textual data.

Causal research collects data that enable decision makers to determine cause-and-effect relationships between two or more variables. Causal research is most appropriate when the research objectives include the need to understand which variables (for example, advertising, number of salespersons, price) affect a dependent variable (sales, customer satisfaction).

Understanding cause–effect relationships among market performance factors enables the decision maker to make “If–then” statements about the variables. For example, as a result of using causal research methods, the owner of a men’s clothing store in Chicago can predict, “If I increase my advertising budget by 15 percent, then overall sales volume should increase by 20 percent.” Causal research designs provide an opportunity to assess and explain causality among market factors. But they often can be complex, expensive, and time-consuming.

Secondary and Primary Data Sources

The sources of data needed to address Research Problems can be classified as either secondary or primary. The sources used depend on two fundamental issues: (1) whether the data already exist, and (2) the extent to which the researcher or decision maker knows the reason(s) why the data were collected. Sources of secondary data include “inside” the company—a company’s data warehouse—or “outside” the company—public libraries and universities, the Internet, or commercial data purchased from firms specializing in providing secondary information. Primary data are collected directly from first-hand sources to address the current information research problem.


When conducting primary research, consideration must be given to the sampling design. When conducting secondary research, the researcher must still determine that the population represented by the secondary data is relevant to the current research problem.

If predictions are to be made about market phenomena, the sample must be representative. Typically, marketing decision makers are most interested in identifying and resolving problems associated with their target markets. Therefore, researchers need to identify the relevant target population. In collecting data, researchers can choose between collecting data from a census or sample. In a census, the researcher attempts to question or observe all the members of a defined target population. For small populations a census may be the best approach.

A second approach, used when the target population is large, involves the selection of a sample from the defined target population. Researchers must use a representative sample of the population if they wish to generalize the findings. To achieve this objective, researchers develop a sampling plan as part of the overall research design. A sampling plan serves as the blueprint for defining the appropriate target population, identifying the possible respondents, establishing the procedures for selecting the sample, and determining the appropriate sample size. Exhibit 2.11 lists the questions and issues researchers typically face when developing a sampling plan.

Sampling plans can be classified into two general types: probability and nonprobability. In probability sampling, each member of the defined target population has a known chance of being selected. Also, probability sampling gives the researcher the opportunity to assess sampling error. In contrast, nonprobability sampling plans cannot measure sampling error and limit the generalizability of the research findings. Qualitative research designs often use small samples, so sample members are usually hand-selected. Sample size affects data quality and generalizability. Researchers must therefore determine how many people to include or how many objects to investigate.



Step 6 is an important step in the research process for descriptive and causal designs. It involves identifying the concepts to study and measuring the variables related to the problem. Given the importance of measurement to the process of creating information, researchers must be able to answer questions such as:

“How should a variable such as customer satisfaction or service quality be defined and measured?”

“Should researchers use single-item or multi-item measures to quantify variables?”

Although most of the activities involved in Step 6 are related to primary research, understanding these activities is important in secondary research as well. For example, when using data mining with database variables, researchers must consider the Measurement Issues. They must understand the measurement approach used in creating the database as well as any measurement biases. Otherwise, secondary data may be misinterpreted.


Designing good questionnaires is difficult. Researchers must select the correct type of questions, consider their sequence and format, and pretest the questionnaire. Pretesting obtains information from people representative of those who will be questioned in the actual survey. In a pretest respondents are asked to complete the questionnaire and comment on issues like clarity of instructions and questions, sequence of the topics and questions, and anything that is potentially difficult or confusing.

If qualitative research is being conducted, interview guides and exercises are developed by researchers. If the data is preexisting “found” data (for example, posted consumer opinions of a book at, then categories or themes are emergent. That is, they are identified by researchers as they review the data. Measurements for observational research may be emergent or predetermined. For example, observers of babies at the Fisher Price toy lab may determine prior to observation that they will count the number of times babies play with each toy as well as the amount of time spent with each toy. Forms will be designed when categories are predetermined to enable researchers to collect the required data.


Data Collection Methods

There are two approaches to gathering data. One is to have interviewers ask questions about variables and market phenomena or to use self-completion questionnaires. The other is to observe individuals or market phenomena. Self-administered surveys, personal interviews, computer simulations, telephone interviews, and focus groups are just some of the tools researchers use to collect data (see Exhibit 2.12).

A major advantage of questioning approaches over observation is they enable the researcher to collect a wider array of data. Questioning approaches can collect information about attitudes, intentions, motivations, and past behavior, which are usually invisible in observational research. In short, questioning approaches can be used to answer not just how persons are behaving, but why persons are behaving as they are.

Observation Method can be characterized as natural or contrived, disguised or undisguised, structured or unstructured, direct or indirect, and human, electronic, or mechanical. For example, researchers might use trained human observers or a variety of mechanical devices such as a video camera, audiometer, eye camera, or pupilometer to record behavior or events.

As technology advances, researchers are moving toward integrating the benefits of technology with existing questioning tools that enable faster data acquisition. Online primary data studies (e-mail surveys, online focus group interviews, Internet surveys) are increasing as well as secondary database research studies.

Preparation of Data

Once primary data are collected, researchers must perform several activities before data analysis. A coding scheme is needed so the data can be entered into computer files. Typically, researchers assign a logical numerical descriptor (code) to all response categories. After the responses are entered, the researcher inspects the computer files to verify they are accurate. The data then must be examined for coding or data-entry errors.

Data preparation in secondary research studies is somewhat different from that used with primary research. Researchers focus on evaluating the use of a single or multiple databases to obtain the needed information. When the data exist in multiple databases, different databases must be merged into a combined database. At times, merging one database with another can be challenging and may require restructuring one or more databases to achieve compatibility. Another activity is determining which data should be included in the analysis.



In Step 9, the researcher analyzes the data and may create summated variables, ratios, constructs, and so on. Analysis procedures vary widely in sophistication and complexity, from simple frequency distributions (percentages) to statistics (mean, median, and mode) and perhaps even multivariate data analysis. Different procedures enable the researcher to statistically test hypotheses for significant differences or correlations among several variables, evaluate data quality, and test models of cause–effect relationships.


Knowledge is created for decision makers in Step 10. Knowledge (as we discussed it earlier in the chapter) is information combined with judgment and interpretation to facilitate accurate decisions. Researchers and decision makers interpret the results of the data analysis. Interpretation is more than a narrative description of the results. It involves integrating several aspects of the findings into conclusions that can be used to answer the research questions.


Step 11 is preparing and presenting the final research report to management. The importance of this step cannot be overstated. There are some sections that should be included in any research report: executive summary/key findings, introduction, problem definition and objectives, methodology, results and findings, and limitations of study. The researcher asks the decision maker whether specific sections need to be included or expanded, such as recommendations for future actions or further information needs. In some cases, the researcher not only submits a written report but also makes an oral presentation of the major findings.