How to Conduct Market Research

Market research consists of a series of steps to collect and analyze market data. Decisions made in one step may affect other subsequent steps, and any modification made in one step often means that other steps may also need to be modified. The steps of market research ① are generally carried out according to the following procedures: ① Determine the questions and hypotheses; ② Determine the information needed; ③ Determine the way to collect the information; ④ Sampling design; ⑤ Data collection; ⑥ Data analysis; ⑦ Survey report.

1. Determine the problem and assumptions

Since the main purpose of market research is to collect and analyze information to help companies make better decisions to reduce decision-making errors, the first step of the survey requires decision makers and investigators to carefully determine and agree on the objectives of the study. As the saying goes, "A proper definition of a problem is half the solution." There are many things that can be investigated in any given problem, and if the problem is not clearly defined, the cost of gathering information may exceed the value of the results presented by the investigation. For example, if a company realizes that its sales have been declining for six months in a row, managers want to know what the real reason is. Is it the recession? Reduced advertising expenditures? A shift in consumer preferences? Or is it poor marketing by the agent? Market researchers should first analyze the information, then identify the research problem and make further assumptions and research objectives. If the researcher believes that the above problem is a change in consumer preference, then he or she should further analyze and make some hypotheses. For example: ①Consumers believe that the company's product design is outdated; ②Competitive product brand advertising design is better.

The main reason for making hypotheses and giving research objectives is to limit the scope of the survey, and to test whether the hypotheses are valid or not from the information derived from future surveys, and to write a survey report.

2. Determine the required information

After determining the questions and hypotheses, the next step is to decide what information to collect, which should naturally be related to the objectives of the survey. For example: ① What is the attitude of consumers towards our products and their brands? ② How do consumers feel about the price of the Company's branded products? (iii) What is the evaluation of the company's brand's TV commercials versus those of competing brands in the minds of consumers? ④ Are there any differences in the attitudes of different social classes toward the Company's brand and competing brands?

3. Determine the way to collect information

The third step requires the development of the most effective way to collect the required information, which needs to be determined: data sources, survey methods, survey instruments, sampling plan and contact methods.

If there is no readily available information (secondary information) that is applicable, the collection of primary information (first-hand information) becomes a necessary step. The method used to collect the information is related to the nature of the information required. It includes experimental, observational and inquiry methods. In the previous example, it was mentioned that the information required is about consumer attitudes, so the market researcher may use the questioning method to collect information. For consumer surveys, it is more appropriate to use personal interviews to facilitate in-depth communication with each other.

4. Sampling design

In the survey design stage should be decided who the sample object is, which raises the question of sampling design. One: whether it is probability sampling or non-probability sampling, depending on the degree of accuracy required by the survey. Probability sampling is more accurate, and can estimate the sampling error, from the statistical efficiency, naturally, probability sampling is better. From an economic point of view, however, non-probability sampling is simple in design and saves time and money. Secondly, one of the issues that must be decided is the number of samples, which in turn requires consideration of statistical and economic efficiency.

5. Data collection

Data collection must be done through the investigator, the quality of the investigator will affect the correctness of the findings. Investigators to the university of marketing, psychology or sociology students is ideal, because they have been trained in survey techniques and theories, can reduce the survey error.

6. Data Analysis

After the data is collected, all answers should be checked and incomplete answers should be considered for elimination or the respondent should be asked again to fill in the gaps.

The data should be analyzed in statistical tables or graphs so that the reader can easily understand the results and see how the statistics relate to the first step of determining the hypothesis of the question. The results should also be presented in the form of percentages and averages for each type of information so that the reader can make a clear comparison of the results of the analysis. However, whether the differences between the percentages and averages of the various types of information are truly statistically significant should be identified using appropriate statistical tests. For example, the monthly consumption expenditures of two types of income households on certain household items may appear to differ on the surface, but whether there is a real difference can be analyzed using the mean test. Information can also be analyzed using correlation analysis, regression analysis and some other statistical methods.

7. Survey Report

The final step of the market research is to prepare a written report. Generally speaking, the written survey report can be divided into two categories: ① specialized report; ② popular report.

Specialized report of the reader is interested in the entire survey design, analytical methods, results and various types of statistical tables, they have an understanding of the technology of market research. The readers of generalized reports are mainly interested in the advice of market research experts. For example, some of the top decision makers in business.