BUSINESS RESEARCH PROJECT

Business Research Project: Questions

Business Research Project: Questions

Answer one

Many researchers are solidly committed to one of two camps, qualitative research or quantitative research. It is observed that they both have their place in formulating a design strategy for large web sites, whose visitor's number in the millions each month.  It is also thought that both interviews and surveys have their place in the arsenal of the design researcher. They are complementary in many design contexts, but rarely can one be substituted for the other without reducing the effectiveness of the results. Although this is a book about customer interviews, some readers will benefit from a brief discussion of when to use these two very common research methods.

Interviews are actually a type of survey methodology, as opposed to an observational method, but in practice the two differ in many respects. They rarely are interchangeable or equally relevant at the same time in the design lifecycle. Surveys are easy to produce and easy to tally, and therefore are attractive to many corporate web site owners. However, surveys have distinct limitations that render their results ineffective for guiding the design of a large web site.

The wrong reason to deploy a survey instead of interviews is: “We have a survey tool, and it will be quicker, easier and cheaper to do a survey than it will be to interview customers in depth.” Many researchers conduct interviews and surveys in cycles that allow them to discover and then quantify the most significant factors.

Surveys do not always produce the kind of data needed to make practical design decisions. Why? Because customers may not know what the options are. Or, they may not be able to understand the way a complex interaction should be designed, just as many people couldn't tell you how their office chair should be designed. The way that interactions are designed on a web site may be very different from the way people think of a task. That is, the interaction model and the mental model are at odds and this is causing customers to drop off in frustration. There isn't any way for a survey to reveal this, because it isn't something users can vocalize or even observe about them. In an interview, the researcher can guide the conversation so that the participant discusses needs and pain points that reveal a solution that they didn't know was possible before the interview, and perhaps even after the interview.

So how do you decide when to conduct interviews and when to conduct surveys?

Surveys are useful when:

It has known the variables that you want to measure

Respondents are able to give you the answers that you see

It is feasible and desirable to reach a representative sample that will respond in numbers that yield statistically significant results

In contrast, interviews are most useful when:

It have yet to determine the most significant factors or variables that are most relevant to a given problem or opportunity

Participants are unable or unwilling to provide you with the answers you seek, but ...