Comprehensive Marketing and Sociological Research

Determine the company's specialization. Don't expect to find a company specializing specifically in your market, such as flooring, the restaurant business, office furniture, children's clothing, etc. These markets are not yet large and structured enough to provide a steady stream of orders that could cover the annual budget of even 2-3 medium-sized research companies. Therefore, industry specialization among research firms has not yet developed. The exceptions, which only prove the rule, are the pharmaceutical, perfume-cosmetic, media, and political markets.
However, specialization by methods and types of research is beginning to form. Some companies specialize only in quantitative methods (surveys, questionnaires, monitoring), while others focus solely on qualitative methods (focus groups, in-depth interviews, expert surveys, content analysis). Some specialize in telephone surveys, known as "call centers."
Companies specializing in qualitative research are generally not large, often comprising 5-7 permanent employees. Companies focusing on quantitative research typically have at least 15 permanent employees unless they are a branch of a large research firm. Large research companies conduct both qualitative and quantitative research. These large firms employ between 40-50 and 120-200 permanent staff, with annual turnovers exceeding 3 million dollars.
If a company specializes in focus groups, it should have the appropriate facilities and equipment (video cameras, VCRs, etc.) necessary for conducting such sessions. They should employ professional moderators (group discussion leaders) and professional recruiters (participants), as well as linguists and psychologists. If a company specializes in mass surveys, it should have field managers, programmers, and analysts on staff. There should also be teams of interviewers and the necessary software.

In a well-developed marketing research plan, control over various sources of error should be included.
General error represents the deviation of the true mean value of the variable of interest in the population from the observed mean value obtained through marketing research. For example, according to the latest U.S. Census results, the average annual income of the population is $75,871, whereas marketing research based on a sample survey reports an average annual income of $67,157.
Sampling error occurs because a specific sample is not fully representative of the population of interest to the marketer. Sampling error is the deviation of the true mean value for the population from the true mean value for the initial sample. For instance, the Census data shows the average annual income as $75,871, while results from a sample survey based on a postal panel (considered quite accurate) indicate an average annual income of $71,382.
Systematic errors are unrelated to sample formation; they can be random or non-random and arise from various causes, including errors in problem definition, approach development, scaling, questionnaire structure, interviewing methods, data preparation, and analysis. Systematic errors consist of non-observation errors and observation errors.
Non-observation error occurs when responses cannot be obtained from some respondents in the sample. The main reasons for non-observation errors are refusals and the respondent being unavailable. Non-observation can lead to the actual sample differing from the initial sample in size or composition. Non-observation error is defined as the deviation of the true mean value of the variable in the original, planned sample from the true mean value in the final, actual sample. For example, the average annual income for the initial sample is $71,382, but only $69,467 for the actual sample, with both figures obtained from a postal panel survey.
Observation error arises when respondents provide inaccurate answers, their answers are incorrectly recorded, or they are misanalyzed. Observation error is defined as the deviation of the true mean value of the variable in the actual sample from the mean value obtained in the marketing research project. For example, the average annual income for the actual sample is $69,467, but according to the marketing research, it is only $67,157. Observation errors may be made by researchers, interviewers, or respondents. Researcher errors include errors in information replacement, measurement, population definition, sampling model, and data processing.
Information replacement error is defined as the deviation of the information needed to address the marketing research problem from the information found by the researcher. For example, instead of obtaining information about consumer choice of a new brand (needed to address the marketing research problem), the researcher obtains information about consumer preferences because it is difficult to obtain information on consumer choice.
Measurement error is defined as the deviation of the information sought by the marketer from the information obtained through the measurement process. For instance, although the marketer is interested in consumer preferences, they use a scale that measures perception more than preferences.
Population definition error arises from the deviation of the actual population size relevant to the problem from the population defined by the researcher. Properly defining the population can be far from trivial, as shown in the example of wealthy households.