Reference no: EM13787373

THE MARKETING RESEARCH PROCESS

Marketing research process
A process used to collect data about marketing programs, external environments, and consumer markets In an attempt to improve the quality of marketing.

The marketing research process is undertaken to answer a wide variety of questions, which might include: "Where do our guests come from? How fre¬quently do people dine out in this area? In what types of restaurants do they most frequently dine? If the seating capacity of a restaurant is expanded by 20 percent, what impact will this have on sales and profits? If the city builds a new convention center, how many additional room nights is that likely to bring to the city?"
Conducting market research is not an inexpensive proposition, and when research is undertaken, care must be taken to ensure that proper methods are used. This is true whether the hospitality organization conducts its own mar¬ket research or relies on external consultants. Market research data are only as good as the methodology used. If poor methodology is used, the results are not likely to describe the situation accurately, and marketing decisions based on this information are not likely to be very appropriate. Figure 6.3 contains the five steps involved in the marketing research process.

FIGURE 6.3 • The marketing research process.

Step 1: Define the Problem
Before initiating any marketing research effort, a firm must decide whether marketing research is necessary.

In general, marketing research should be undertaken if it clarifies a problem that could impact your business, if it helps in selecting between alternatives for achieving marketing objectives, if it assists in gaining a competitive advantage, or if it provides useful information on your markets.1 Marketing research may not be needed if the information is already available, there is insufficient time for marketing research, resources are not available, or costs of conducting the research outweigh the potential benefits of having the information.
If the decision is made to proceed with the marketing research, the research problem should be clearly defined. What does the research effort propose to do? What types of questions need to be asked? What solutions are sought? A strong tendency among all researchers, especially novice researchers, is to rush into data collection without giving adequate thought to defining the problem. This tendency should be vigorously avoided. A small amount of time spent in defining and refining the problem will save many hours later on.
First, it is necessary to view the problem from the marketing manager's perspective. Normally, a problem is brought to the attention of a manager when there is a decrease in a performance measure such as sales volume, profit, or market share. It is important not to mistake this for the problem; instead, recognize it as a symptom of an underlying problem. For example, if a hotel's occupancy rate has suddenly fallen, there could be a number of causes. A new competitor may have opened, current competitors may be discounting prices or increasing advertising expenditures, consumers may be dissatisfied with the hotel, there may be construction in the immediate area, or there may be a downturn in the economy. Any one of these factors, or a combination of them, could be the cause of the decrease in performance.
Once the marketing manager has pinpointed the problem, it is time to formulate the marketing research problem. In other words, what informa¬tion does the marketing manager need to solve the problem? This involves taking constructs (marketing terms or concepts) such as brand loyalty and customer satisfaction and establishing an operational definition that can be used to measure these constructs. For example, brand loyalty could be measured by asking travelers for the percentage of their flights that are on a particular airline. The researcher must also identify the relationships between constructs. One could assume that as customer satisfaction increases, so does brand loyalty. There may even be prior research to support this relationship.

Research design
A master plan specifying the methods and procedures for collecting and analyzing the needed information.

Exploratory research
This is research used to determine the general nature of the problem.

Descriptive research
This is research that helps answer the questions who, what, where, when, why, and how.

Causal research
This is research used to define cause-and-effect relationships between variables.

Secondary data analysis The process of reviewing existing information that is related to the research problem.

Focus group
A group of 8 to 12 people who represent the population being studied and are brought together in an informal setting to discuss the issues surrounding the research problem.

Step 2: Plan the Research

The second step in the marketing research process involves planning the research design for obtaining the desired information to address the research problem. The research design is basically a master plan specifying the meth¬ods and procedures for collecting and analyzing the needed information. There are three main categories of research designs from which to choose, based on the objectives of the research exploratory research, descriptive research, and causal research.
EXPLORATORY RESEARCH. This type of research is used to gain background information when a firm doesn't have a good understanding of the nature of the problem. Exploratory research can be used to obtain additional informa¬tion about a topic and to generate hypotheses stating the relationships between two or more variables. This research tends to be informal and unstructured, and it is mainly used to gain background information, define terms, and clar¬ify problems. The hypotheses that are generated using exploratory research can be tested in future research efforts.
Some of the more common methods for conducting exploratory research are secondary data analysis, experience surveys, case analysis, and focus groups. Secondary data analysis is the process of reviewing existing information that is related to the research problem.

The sources for secondary data analysis were discussed previously in this chapter, along with the advantages and dis-advantages associated with using this method.
Experience surveys are similar to the qualitative methods discussed in the sales forecasting section of Chapter 5. Basically, information is gathered from individuals who are believed to be knowledgeable about the research topic. Case analysis refers to the use of information about a situation that resembles the current situation surrounding the current research problem. Hospitality firms could benefit from the experiences of other firms in their industry, or of firms in other industries that faced similar circumstances.
One of the most common methods of exploratory research is the focus group. Focus groups consist of 8 to 12 people who represent the population being studied and are brought together in an informal setting to discuss the issues surrounding the research problem. The sessions generally last from one to two hours and are guided by a moderator, who ensures that all the group's members give input and that all of the pertinent topics are covered. Focus groups are valuable for testing new product designs and service concepts, test¬ing advertising campaigns, and gaining insight into the market's basic needs and attitudes. The use of focus groups has gained in popularity because the respondents can interact as the research client watches through a two-way mirror. Many focus group sessions are videotaped so they can be examined in

more detail at a later date by a number of different people. The potential weaknesses of focus groups are that, given the small number of participants, the group may not completely represent the population of interest, and, because the groups are unstructured, the information provided during the ses¬sion is often very subjective and open to interpretation. Finally, it may be dif¬ficult and expensive to obtain the participants, the moderator, and a facility for a focus group.
DESCRIPTIVE RESEARCH. The second category of research design is con¬cerned with answering the basic questions of who buys the product, what cus¬tomers do with the product, where they buy, when they buy, why they buy, and how they buy. Primarily, the researcher is trying to profile the customer base in terms of demographics, psychographics, attitudes, and purchasing behavior.
A cross-sectional study measures the population at one point in time; it provides a snapshot of the population. This type of study is normally used to address a particular problem when it arises. For example, one hotel sur¬veyed customers in its atrium restaurant over a one-week period to obtain information that could be used to redesign the restaurant and its menu. A longitudinal study is used to measure the same population over an extended period of time. Generally, these studies use the same sample, referred to as a panel, and collect the same information over time (e.g., once a year). Lon¬gitudinal data are useful in depicting trends and changes in consumer needs and attitudes. For example, many firms offering syndicated services will track the trends in an industry by constructing and maintaining a panel of con¬sumers who complete an annual survey.
CAUSAL RESEARCH. The third category of research design focuses on cause-and-effect relationships that are pertinent to a research problem. A series of if-then statements can be used to model certain elements of the hospitality ser¬vice. For example, a hospitality firm may want to examine how a particular change in the marketing mix affects sales, market share, and/or customer satisfaction. The following relationships could be tested. If the quality of food is improved, customer satisfaction will increase. If on-time performance improves, market share will increase. If an additional salesperson is hired, occupancy rate will increase. The potential benefits of understanding causal relationships are great. Firms could design better products, create effective advertising campaigns, and set prices that will maximize revenue. Unfortu¬nately, many factors affect the consumer decision-making process, either alone or in combination, and it is almost impossible to understand them completely. In fact, many researchers argue that there is no such thing as true causality. Regardless, as managers' understanding of consumer behavior increases, the firm's performance will improve.

Cross-sectional study
A study used to measure the population of interest at one point in time.

Longitudinal study
A study used to measure the same population over an extended period of time.

Step 3: Collect the Data

Observation
A process involving watching consumers and documenting their behavior.

Three major decisions must be made during this stage in the marketing research process. They involve choosing a data collection method, designing data collection forms, and determining the sampling plan. Weaknesses in any of these areas can have a profound impact on the reliability and validity of the results. As previously noted, reliability refers to the consistency of responses to questions, and validity refers to the accuracy of the measure. It is possible to have reliability without validity, but for a measure to be valid, it must also be reliable. For example, a poorly written question can solicit consistent responses, but it may not be a valid measure of the construct that is being studied. Even with the best sample, a poorly designed collection form will result in inaccurate data. Similarly, a well-designed collection form is useless if the sample does not represent the population.
DATA COLLECTION METHODS. Once the research design has been planned, it is necessary to determine the best method for collecting the data.

The three possible methods for obtaining primary data are observation, surveys, and experiments. The appropriate method will depend on the research objectives and the research design. Exploratory research is most often accomplished using observation and surveys. Descriptive research uses all of the data col¬lection methods, but it relies very heavily on information received from sur¬veys. Finally, experiments are used almost exclusively for causal research. However, observation and surveys are also used to obtain information to study causal relationships.
Observation involves watching consumers and making organized notes to document or record the observed behavior. When doing this type of research, it is important that all individuals acting as observers record their observations in the same manner. Direct observation refers to the process of observing behavior as it actually occurs, by either a person or a mechanical device. For example, hotels and restaurants use "mystery shoppers" to expe¬rience the service firsthand and record their observations. Similarly, some amusement park chains have cameras and employees that are used to observe the behavior of the employees to make sure they are paying attention and adhering to safety procedures. Indirect observation refers to the process of observing behavior after the fact. For example, restaurants and airlines can search the trash to determine the eating behaviors of their guests and pas¬sengers. This physical trace evidence is useful in deciding which meal items to include and which ones to eliminate. In most cases, the observation is dis¬guised so as not to affect the consumer's behavior; however, there are instances, such as food testing, when subjects know they are being observed.

Unfortunately, subjects' behavior can be difficult to interpret, especially the attitudes and motivation behind it.
Surveys involve asking consumers to provide information regarding the issues surrounding the research problem on a questionnaire or comment card. The survey can be filled in by the researcher, completed with the aid of a com¬puter, or completed by the respondent (self-administered). When used prop¬erly, the survey method can gather a great deal of useful information. The survey method is adaptable to a variety of situations and is relatively inex¬pensive. Surveys may be accomplished using a number of different methods, including telephone surveys, direct mail surveys, or personal interviews. A summary of the advantages and disadvantages of these methods is provided in Table 6.2.
Telephone surveys are the most common method of survey data collec¬tion because they are the easiest to implement and produce very quick results. Another major advantage of this type of survey is the cost. No travel is involved, and a single individual may contact and solicit answers from a large number of people in a fairly short period of time. However, there is no face-to-face contact, and people are often not inclined to answer ques¬tions over the phone, especially if they are complicated. Therefore, the reli¬ability of the answers received over the telephone can be an issue. Another

Surveys
Data collection instruments designed to gather specific information for a particular research problem through a series of questions and statements.

example is the national "do not call" list that prohibits telemarketers from contacting individuals who chose to be on the list, which can lead to a biased sample. However, political polls and noncommercial surveys are not included in the prohibition.
Direct mail surveys offer ease of completion, respondent anonymity, and a low cost per response. However, there are a few major drawbacks. First, the response rate is normally quite low, and the collection process is slow. Often, less than 25 percent of the surveys are properly completed and returned, and it may take up to three months, and two or three mailings, to obtain an ade¬quate sample. With such a low response rate, there is a risk that the individ¬uals who returned the surveys do not represent the population, and so any decisions based on the results could be biased. Second, direct mail surveys do not allow any in-depth questioning, and they do not allow for follow-up ques¬tions. The respondent merely sees the written questions and has no opportu¬nity for clarification. This may make it more difficult to generate answers that reflect the complexity of opinion within the targeted market segments.
Personal interviews allow more in-depth questioning. An interviewer nor¬mally uses a guide sheet to direct the interview and may adjust the question¬ing to focus on a point of special interest or to follow up an answer given by the respondent. There are two drawbacks to personal interviews as a survey¬ing technique. First, the major drawback is cost per interview. It is extremely expensive to have an interviewer spend a long period of time with each indi¬vidual in order to gather information. An in-depth interview can last as long as an hour; hence, the number of individuals who can be interviewed is lim¬ited, and the cost per interview is quite high. The cost of travel also makes this type of survey expensive. Second, a good deal of training must be done for interviewers to be effective. In addition, supervision is required in order to have control over the interviewers. Intercept interviews are a form of personal inter¬view conducted in major traffic areas such as shopping malls, in an attempt to eliminate some of the drawbacks associated with cost and speed of response.
Electronic surveys are increasing in popularity because of the availability of email address lists and advances in technology. There are two basic types of electronic surveys: web-based surveys and email surveys. Web-based sur¬veys typically provide immediate feedback for the researcher, and simplify the data coding and editing process. Email surveys are either embedded within the email message, or attached to message. They must be completed and returned to the researcher, and require data input and editing. Also, links to Web-based surveys can be contained within an e-mail message. The major advantages of electronic surveys are the low cost per respondent to collect data (less than direct mail), the speed of response (most responses are collect within two weeks), and the ease of data input (if Web-based). The major drawbacks

of this approach are that e-mail addresses can be outdated and often filter incoming mail, and the response rate is low because of all the clutter and spam mail. However, this method combines the advantages of direct mail and tele¬phone surveys. Popular Web-based survey programs such as SurveyMonkey are readily available to any size organization or individual, and the programs are easy to use.
A comment card is a particular form of survey that is frequently used in the hospitality and tourism industry. The main objective of the comment card is to gather information from customers about their lodging or dining experience to determine if they are satisfied with the service. The card is normally placed in the guest's room in a hotel or on the table in a restaurant. This allows the cus¬tomer to provide immediate feedback regarding the service and its delivery. In addition to immediate feedback, comment cards also have other benefits: (1) they are less expensive than other survey methods, (2) they can be tailored to the needs of the organization, (3) they can be standardized for easy analysis and comparisons, and (4) customer needs can be tracked over time. Figures 6.4a and 6.4b are examples of typical restaurant and hotel comment cards.
There are also some disadvantages associated with comment cards: (1) there is often a low response rate and the sample may not be representative of the population, (2) they can deal with only a limited number of issues because of their short length, and (3) there can be problems with reliability

and validity. However, it is possible to improve the response rate by handing them out personally (e.g., waitstaff or front desk staff), keeping them simple (e.g., using closed-ended questions), leaving a space for comments, offering an incentive, and/or promising confidentiality (e.g., use a collection box or postage-paid return).
When conducting experiments, a researcher divides the sample of people into groups and exposes each group to a different treatment while trying to control for other extraneous factors that may affect the outcome. The treat¬ment variable is referred to as the independent variable, and the outcome of the treatment is measured using a dependent variable because changes in the variable are dependent on changes in the treatment. In other words, there is a cause-and-effect relationship in which the independent variable is the cause and the dependent variable is the effect. Experiments can be conducted in the field under normal conditions or in a laboratory setting, where extraneous fac¬tors can be more easily controlled. Marriott International has been experi¬menting with its sales force in recent years to find the best approach. Recently, the hotel chain has allowed some salespeople to work from home and sell mul¬tiple properties rather than sell one property from an office in the hotel. The performance of these two groups in terms of volume and revenues can then be compared to determine the best approach.
Test marketing is a common form of field experiment consisting of manip¬ulations in the marketing mix at certain locations that represent the compet¬itive environment and consumer profile of the overall population. For exam¬ple, all of the national fast-food companies use cities across the country as test market centers. In each city, the companies will introduce or test-market new products or marketing mix changes to obtain customer reactions and to proj¬ect future sales. Sales may then be compared with those at other test market centers to determine the popularity of new and old items and to decide which products will be introduced systemwide. It is quite expensive to conduct this type of study, and it is also quite difficult to control all external variables that may affect the outcome of the experiment. For instance, extraneous variables such as the weather or the advertising efforts of competitive hospitality oper¬ations could easily have an effect on the sales volume of new products. Test markets are common for new restaurant concepts as well. Darden Restau¬rants recently opened a new concept called Seasons 52 near its headquarters in Orlando, Florida. The marketing mix and menu were tested and the new chain was expanded to other areas in Florida and Atlanta, Georgia, once the company was comfortable with the business model.
To ensure good flow and a high response rate, it is necessary for researchers to follow some basic guidelines in organizing questionnaires. The following format is recommended to improve the probability that respondents

Experiments
A data collection process used to compare a control group with one or more treatment groups to determine if there are any differences attributed to the variable(s) being tested.

Test marketing
A common form of field experiment consisting of manipulations in the marketing mix at certain locations that represent the competitive environment and consumer profile of the overall population.

Population
The entire group, or target market, that is being studied for the purpose of answering the research questions.

Sample
The subset of the population that is drawn in such a way so as to represent the overall population.

Sampling unit
The basic level of investigation in a research study.

Census
A sample consisting of the entire population.

Sampling error
The difference between the sample results and actual population measures.

will participate in the survey and complete the questionnaire. First, the ques¬tionnaire should start with some easy questions to screen respondents based on their purchasing behaviors. Respondents who fit the desired profile pro¬ceed to the next section, and those who do not are thanked for their time and cut loose. Second, more detailed information is obtained regarding respon¬dents' purchase and usage behaviors so they can be categorized. This section focuses on basic information that is easy to provide, and it serves to warm up the respondent for later sections. Third, respondents are asked to answer a series of complicated questions, including ratings and rankings. The fourth and final section of the questionnaire is used to obtain background informa¬tion on the respondent. This demographic and psychographic information is used to classify individuals for statistical analysis.
The goal of questionnaire organization is to select respondents who rep¬resent the target population and ease them into the questionnaire so they become involved. Then they are less likely to cut off in the middle of the ques¬tionnaire when they are confronted with more difficult questions and sensi¬tive questions dealing with topics such as age and income. Therefore, it is important to pretest the questionnaire on a small sample to check the ques-tion wording and organization. Any problems can be identified and corrected before attempting to use the data collection form on the final sample.
DETERMINING THE SAMPLING PLAN. Sampling is the process of using a small subset of the population to obtain information that can be used to make inferences about the total population. A population is the entire group, or target market, that is being studied for the purpose of answering the research questions. A sample is the subset of the population that is drawn in such a way so as to represent the overall population. A sampling unit is the basic level of investigation. The sample unit could be an individual, a household, or a firm or organization. For example, studies in the hospitality industry could look at a hotel's guest, a specific hotel property, or a hotel chain. Normally, the cost of a census, or the investigation of the entire population, is prohibitive, and the survey would take too much time to complete. There¬fore, a sample is used in hopes that the results can be applied to the overall population. Whenever a sample is used instead of a census, there is some difference between the sample results and actual population measures. This difference is referred to as sampling error.
The first requirement in sampling is to define the population from which to gather information. The population is simply a definition of the group of individuals from which to gather information. For example, two specific populations might be (1) all males and females between the ages of 20 and 26 who are not married, and (2) all males and females who earn more than $25,000 per year and work within one mile of the restaurant.

The next step is to determine the number of individuals, or sampling units, to survey. This is known as the sample size. The determination of sample size can be based on financial, statistical, and/or managerial issues. From a finan¬cial standpoint, the size of the sample depends on the available budget, or what the client can afford. From a managerial standpoint, the size of the sam¬ple can be a function of an arbitrary judgment or based on similar studies. From a statistical standpoint, the size of the sample may depend on the type of analysis that will be used. In general, more sophisticated statistical tech¬niques tend to require a certain number of observations to be valid.

Sampling units
The chosen number of sample units to be included in a research study.

Step 4: Analyze the Data

Two basic forms of statistical analysis are used in marketing research: descrip¬tive analysis and inferential analysis. Descriptive analysis uses aggregate data to describe the "average" or "typical" respondent, and to what degree respon¬dents vary from this profile. The measures used for central tendency are the mode, median, and mean. Mode refers to the value that occurs most often. Median refers to the value that represents the middle of an ordered set of responses. In other words, the responses are ordered from high to low, or low to high, and then the middle value is determined. Mean refers to the arith¬metic average, or the sum of all responses divided by the number of responses. The measures of variability are the frequency distribution, range, and stan¬dard deviation. The frequency distribution provides the counts for each value in the set of responses. The range is calculated by taking the difference between the highest value and the lowest value of the ordered set of responses. The standard deviation is a measure of variance between the observed values and the mean for the set of responses.
The second form of statistical analysis is inferential analysis, which is used to test hypotheses and estimate population parameters using sample statistics. Statistics such as the t-statistic and the z-statistic are used to test for differ¬ences between the sample mean and a hypothesized mean. These test statis¬tics can also be used to test for differences between two groups, and analysis of variance (ANOVA) is used to test for differences between more than two groups based on their respective means and variances. Correlations and cross-tabulations are used to determine if an association exists between two vari¬ables. If so, the two variables will vary together either directly or inversely. Finally, multivariate statistics can be used to test for relationships between more than two variables. These forms of statistical analysis are beyond the scope of this text, and interested readers should refer to a marketing research text for more details.2

Descriptive analysis
An analysis using aggregate data to describe the "average" or "typical" respondent, and to what degree respondents vary from this profile.

Inferential analysis
An analysis of cause-and-effect relationships used to test hypotheses.

Step 5: Prepare the Final Report

Once the research is completed and the data are analyzed, it is necessary to prepare a final report that provides a detailed outline of the research design, summarizes the results, and provides some conclusions or recommendations. The researcher should consider the audience for the presentation when preparing the final report. Both written and oral reports have been criticized for things such as excessive length, impractical recommendations, and the use of complex terms. These mistakes can be avoided if the researcher determines the personality and requirements of the audience and takes them into con¬sideration when preparing the report. For example, many clients prefer to be shown the results summarized in tables and charts rather than to read detailed discussions including statistics.
There are some guidelines that can be followed when preparing the final report that will improve the probability of client satisfaction, and success. First, the research team should plan to devote an adequate amount of time to preparing the report. In fact, the time for report preparation should be included in the time frame outlined in the proposal. Second, the original proposal should be examined and the research objectives should be addressed in the final report. Third, the research team must understand the needs of the audience and determine the content and length that are appropriate for the report. Fourth, it is important to antic¬ipate possible objections or concerns and to address them in the report or presentation.
Most written reports follow a standard outline. The report normally begins with an executive or management summary that clearly and con¬cisely states the project's objectives, methods, conclusions, and recom¬mendations. Next, the actual body of the report begins with a detailed discussion of the research objectives, followed by an explanation of the research methodology, including its advantages and limitations. The research methodology section contains the elements of the research plan, including questionnaire design, sampling, data collection, and type of analysis. The next section contains a detailed description of the results, with references to charts, figures, and tables that summarize the results. Finally, the report ends with the conclusions, implications, and recom¬mendations of the research team. Any tables, charts, figures, or other supplemental materials will appear in an appendix at the end of the report. Examples of supplemental materials would be an annotated questionnaire containing the results for each question, or a list of responses to open-ended questions.