Wednesday, August 24, 2005

Put the Focus Back Into Focus Groups

At the recent AAAA account planners' conference (AdAge, 8/8/05), Malcolm Gladwell (Blink, Tipping Point author) threw down the gauntlet on decades of market research tradition with the assertion focus groups should be banned. This is akin to saying email should be banned because much of it is poorly written. Mr. Gladwell’s view of focus groups as random groups of people asked to make decisions on marketing in an artificial setting bears little resemblance to the reality of qualitative research today. After conducting hundreds of groups in the past two years, our conclusion is that it’s not the technique that is flawed, but how it’s used.

Focus groups are fast and easy to execute. The flexible format makes them ideal for exploring strategic issues (who is my target?) to tactical decisions (is a red arrow more readable than a copy block?). Before the M&M’s get thrown out with the candy dish, we think it’s worth trying to putting the focus back into focus groups. Here are some suggestions for making qualitative research more meaningful and less prone to misleading errors.

When appropriate, conduct groups online. A series of groups comprised of people from all over the country is automatically more representative than groups conducted in Denver, Chicago and LA. Our experience is that online groups are also better recruited and less susceptible to manipulation by individual clients or respondents. Here’s why:

Setting: Less artificial as respondents are in their own home, office or library rather than a sterile focus group facility. Observing from any Internet enabled location increases client participation.

Sample: Even hard to find populations can be identified and recruited. There are fewer ‘professional’ respondents since recruiting is not limited to those living near a facility. We have successfully recruited owners of $2000 exercise machines, users of flavoried rum, even concession stand volunteers.

Control. Loud-mouths have a harder time dominating in a forum where everyone speaks at the same time and at the same volume. Less time is spent controlling and more on listening.

Validity. People are more likely to admit to habits or opinions that they think others would disapprove of in the anonymity of cyberspace than they would face-to-face.

Add a quantitative component. A screener is merely a quantitative survey in sheep’s clothing. It can be treated as part of the overall research design in order to put qualitative findings in a larger context. Such “hybrid’ designs’ reduce overall costs by sharing the burden of recruiting across two studies. For example, in a recent study a survey was used to identify cross channel shopping habits for a large retailer, focus group respondents were selected based on their survey responses. The resulting samples were unusually homogeneous, allowing nuances in behavior and attitudes to be identified.

Make focus groups more observational. Groups where the moderator asks a series of yes no questions is nothing more than small sample quantitative survey. Effort and creativity are required to encourage respondents to talk to each other so their natural responses can be observed. Interactive collaborative tools -- such as trips to the web, projective techniques, survivor-type elimination games and whiteboard exercises -- make the experience fun and involving, as well as reduce artificiality.