A company CEO believes strongly that his reinsurance company can successfully enter the auto market by promising dealers a 30 percent increase in gross profits for zero investment. He wants to develop the branding around this theme — a 90-degree change from the business’s current position. Unfortunately, the board sees the shift as too risky. Currently, there’s a stalemate disrupting progress, and the only way to break the logjam is to find out whether the CEO’s instincts are on the money. The immediate questions that validate a “success-conclusion” are:
- Is the promise realistic? Even if so, it’s only half the conversation.
- Is it significantly better than what competitors can offer?
This mini case study shows us that theorizing is one thing, while proving something is quite another. To get the rest of the team behind the idea, the CEO must conduct a systematic and objective research effort. Indeed, everything he does from that point depends on the research design to rubber-stamp investment in the new initiative.
What is research design?
Research design aims to methodically resolve a research problem by structuring a roadmap to collect, measure, and analyze relevant data. A research problem is anything in the market that’s puzzling the stakeholders in a business who are trying to advance a company’s competitive positioning. Companies rely on research design to reveal obstructions, opportunities, or both. Alternatively, management can prove a theory they think is valid with verified data and assessment, as demonstrated by the reinsurance proposition above.
Research routes can go in a quantitative or a qualitative direction. There is, of course, a middle road approach — a hybrid option, if you will — integrating the two types of research design to get the most out of collected data. Nonetheless, the research mechanics are substantially different, so let’s get into the nitty-gritty of research design.
Quantitative research design
Quantitative research design relies on substantial sample sizes, focusing on the response volume which is then broken down numerically and filtered through statistical analysis. Thus, it examines how respondents react to the same stimuli, not why. In other words, it doesn’t get into the emotional and cognitive drivers behind the responses. The process keeps things standardized and on a level playing field by asking respondents to answer the same questions. As a result, research uniformity eliminates biases, but there are exceptions. Variations may arise from the way the company decides to conduct interviews, the most common being:
- Online surveys
There is a place for creativity in quantitative design. When relevant, particularly in surveys, respondents may reach a point where they can choose between A or B for a specific question. The following question may be: “If you answered A to the question above, please answer the following.” Another option would be to allow “Other” responses within a quantitative research project, along with a box that asks the respondent to explain his or her answer.
Mostly, however, researchers don’t encourage open-ended responses. Why? Because they dilute the researchers’ ability to quantify results. The more open-ended it becomes, the more it moves to the qualitative side of the research design spectrum. Conversely, closed-ended questions create numerical efficiency, although it sacrifices the ability to dig below the surface into motivations. In addition, coding quantitative answers takes less time, eliminating the burden of categorizing opinions that can differ widely.
Qualitative research design
Qualitative research design focuses on emotional drivers and thoughts that underlie customers’ market behavior. It aligns closely with behavioral and psychographic segmentation. Moreover, it frequently requires the services of trained interviewers with a psychology or sociology background to draw conclusions from observing respondents via video, audio, group discussions, and text in emails and mobile SMS. Relative to quantitative research, it’s a substantially more expensive format — restricted only to a handful of respondents (versus quantitative’s collection of hundreds, even thousands, of responses in much less time.)
The open-ended questionnaire is fundamental to this type of design research, probing to uncover the “how” and “why” of market behavior. For example, “Why did you switch brands after ten years of loyalty?” or “What are the most compelling reasons motivating you to buy more?”
Bias in this type of research
Bias in qualitative research boils down to interviewees assuming the role of expert or opinion leader while being interviewed in an unnatural setting. In other words, what people say in one situation is quite different from how they feel or think about the same thing when an actual buying decision occurs. For this reason, specialized training and expertise are essential to draw accurate conclusions.
Analyzing the data
Qualitative data analysis breaks down into these five categories:
- Content analysis: Systematically compartmentalizing verbal and observed activity data into a format that anyone can understand. It involves skills of classifying, summarizing, and tabulating.
- Narrative assessment: Reformulating respondents’ often diverse stories to extract the primary qualitative data related to the company’s marketing performance.
- Discourse analysis: Interpreting text or recordings of natural conversations in different situations.
- Framework analysis: A process of taking the data through stages that begin with familiarization to identify a thematic framework, followed by coding, charting, and mapping out the data, and finally, interpreting what it all means.
- Grounded analysis: Often connected to thematic theory, begins with analyzing one observed situation to develop a thesis, and looking at a broader landscape to see if other cases support the original proposition.
Qualitative research methods
Researchers apply techniques like in-depth unstructured interviews, focus groups, case study research, and ethnography to get to the crux of how respondents think or feel about things. It can spell the difference between marketing success or failure. Interviews, simple “sit-back-and-observe” formats, cultural and sociological records, and personal intuition all come into play with qualitative research. The researcher coaxes the emergence of valuable information by creatively providing direction and confirming theory validity.
Hybrid research design: the best of both worlds
Quantitative research responses sometimes rely on the question, “Why did you select that response?” However, in most cases, that’s a question that only qualitative research design can adequately address. An integrated research program that seamlessly melds quantitative and qualitative design can deliver insightful research. Therefore, marketers who seek the most accurate market segmentation information will often rely on a combination of quantitative and qualitative research designs. That means relying on professionals in the social sciences to help zero-in on respondents’ motivational drivers, but it’s certainly worth it to take this extra step.