Part Two: Marketing Research Implementation
In Part One of the Marketing Research Process, we discussed the first three steps of the research process:
- Determining the Need for Marketing Research
- Defining the Research Problem
- Setting the Research Objectives
We characterized those steps as “Thinking About the Research”. Once you’ve completed the thinking, you have created the blueprint for the information development. The next six steps of the process involve decisions on methodology, data collection, sampling, analysis and reporting.
Step Four: Determine the Research Design
In Step Four, we identify the optimal way to collect the information needed to meet the research objectives. Again, there are a series of questions that researchers ask in developing the optimal research design:
- Is the information we need to make this decision already available? You may have internal information that can help you make this decision. Or there may be industry, trade association, foundation or academic research information sources that you can use to find the information you need. Finally, there are also many government information sources that can help you as well.
|Information that has been collected for a different purpose than the one at hand is called secondary research. While it may not be perfect (it may be old, or not in quite the right format) secondary information is usually very cost effective and should be your first line in answering business questions. Information that is collected specifically to address the issue at hand is known as primary marketing research.|
- Do I know enough about how my respondents think and speak about the topic to design the research appropriately? In addition to secondary research, you may need a phase of exploratory research to better understand your respondents, and how the survey instrument (aka the “questionnaire”) should be structured and the questions worded. In marketing research, exploratory research is often done using qualitative research techniques.
|Qualitative research uses unstructured or semi-structured question techniques to allow a good exploration of the topic at hand. The sample size is typically small, and respondents are selected to fulfill a given quota. Examples of qualitative research include in-person and online focus group discussions and in-depth interviews. Quantitative research generates numeric data or data that can be transformed into statistics. Quantitative research usually includes larger sample sizes with the intention of generalizing the findings to the population of interest. Quantitative research is usually conducted through surveys, with online and telephone surveys being used most frequently.|
- How will I analyze the resulting data? Understanding any special analytical techniques that will be used can influence your choice of methodology. If you are planning on using a technique such as conjoint analysis or Max Diff analysis, there are specific guidelines to study design and you need to pay special attention to survey length to avoid respondent fatigue or boredom, which could jeopardize survey quality. If you are asking respondents to choose between multiple options, you will want to consider the maximum number of choices that you can show to your respondents without introducing a bias, and choose your methodology to allow you this control.
- How can I best reach my respondents? Do your respondents have access to the Internet? If so, then an online survey may be most appropriate. If not, then a telephone survey may be desirable, in which case, you need to ascertain whether or not you need to contact people on their landline, cell phone, or both. Do you have to share complex information for the respondent to read or pictures or video that must be reviewed? In this case, an online survey will provide the best solution.
Step Five: Design the Questionnaire
Finally! We are now ready to design the questionnaire. One of the best ways to design an efficient survey is starting with the end in mind. First write an outline of the report that you plan to provide; outlining what the potential answers are to your research objectives. This will indicate the questions you need to ask to get the information needed to fill in the report. Now is also a good time to go back to steps two and three and make sure that your questions fit the Problem Definition and Research Objectives. Because writing a good, unbiased questionnaire is as much art as science, we will consider this topic in a separate blog post in this series.
Step Six: Determine the Sample Frame and Sample Size
The sample frame refers to the source or segment of the population from which you will draw the sample for your questionnaire. Who do you need to reach to get the information you need? Are they individuals or businesses? Are they defined by demographic characteristics, product usage, customer experiences, or something else? What is the incidence of the respondents in your sample frame (the percent of the population they represent)? Do you have access to lists of respondents in your sample frame that include contact information (email address, telephone number, etc.) or do you need to purchase a sample from a panel provider? There are many, many ways to identify the sample frame and source the sample, and these will be discussed in a future blog post.
Similarly with the sample size: there are many considerations to determine the appropriate sample size. In general, sample size has a direct effect on reliability; the bigger the sample, the greater the reliability of the results. But, sample size also has a direct effect on cost. So, balancing reliability with budget is the key in determining the optimal sample size. You need to not only have a good idea of your budget, but also need to know what your tolerance is for variance in the data caused by sample error. We will also address sample size and sample reliability more specifically in a future blog post.
Step Seven: Collect the Data
Sound data collection practices are important to minimize bias and error in the data. Additionally, data collection techniques and management can be critical to achieving an acceptable response rate from your respondents. Data collection procedures must be consistent and controlled. If you are using telephone interviewers, they all need to have the same training and instructions and must read each question exactly as it is written with no deviations or omissions. For online surveys, the program must be checked to be sure that skip patterns and branching correctly display appropriate questions for respondents.
Step Eight: Data Preparation and Analysis
Most survey techniques allow data to be automatically entered into a database. However, some paper-based methodologies will need data entry as the first step. Once the data is downloaded to a database, the researcher needs to inspect the data to ensure it is error-free prior to any analysis or interpretation of the data. Open-end questions are coded into categories to facilitate analysis and interpretation.
Once the data is thoroughly prepared, basic analysis (called frequencies, ordinals, or counts) give the researcher a first look at the distribution of responses for each question. Crosstab analysis (using some questions to slice the data into subsets) is frequently used in marketing research to identify differences between subgroups. Finally, advanced statistical analysis techniques (such as multiple regression, conjoint or discrete choice analysis, factor analysis, and segmentation analysis, etc.) can be used for additional insight.
Step Nine: Develop the Report and Present the Findings
The final step in the research process brings the work full-circle, as research findings are organized to address the research problem and objectives defined in Steps One and Two. Research reports and presentations should focus on addressing the questions being asked by the audience – that should have been specified and have formed the basis for all the decisions made about the research project. Again, due to the critical importance of this topic, we will discuss this further in a future blog post.
Using the Nine-Step Research Process may appear to take a long time, but it only adds a few hours to the project timeline. And those few hours can mean the difference between conducting useful, impactful, and important research or wasting time and money on research that does not address the business questions. Take control of your marketing research by implementing the discipline of the Marketing Research Process.