Groups Vs. One-On-Ones

One of the things that has always given me pause for thought is whether to use groups or one-on-one interviews to conduct qualitative research. I've finally figured it out. Focus groups are generally good at producing middle-of-the-road results - say the kind of results you want when you want to reach an audience in a very general way - like with mass advertising. In this instance someone is likely looking for a lowest-common-denominator type of marketing. I will clarify that there are a lot of uses for this kind of approach. One-on-one interviews, however, are much more useful when testing something that is new, unique or takes time to be adopted.

Let's start with groups. Early on in my career, someone asked me "why do groups when you can just do a series of one-on-one interviews?" The answer always stuck with me - focus groups produce "group dynamics", and those are the key results of focus groups - it is not necessarily what people say that is important. We get to watch how an idea changes, where there is resistance to it, and where there is acceptance of it. We get to see how strongly people hold on to ideas, and how they react to having their ideas changed, rejected or challenged. A marketer can observe these dynamics and figure-out what arguments or factors will move people and change their opinions. This is why I love using groups to evaluate policy issues, which are very dynamic and flexible.

What prompted me to write this blog is a continual interest in why groups can still perform so badly at predicting certain product successes/failures, and why people always assume that "mediocre" products are something that have been "focus grouped." I finally came up with my solution. In observing group dynamics, I've come to realize that the group is always moving towards (or away from) something. Ten people take an idea, play with it and either change it or reach an opinion about it. When group processes like this happen, you can't help but get a watered-down version of the original.

In a focus group people are not themselves - it is sort of like "mob mentality" minus the violence - we would do and say things in a group that we would not do and say as individuals (hmmmm - maybe that's why groups are such bad predictors of behaviour). In a focus group, people ARE swayed by dominators - they are shy and they do advance their agendas. A moderator is there to observe, balance and interpret these phenomena. As such, they are not negatives - this is what happens in real life as people live in a dynamic world, and if what is being tested is something that needs to meet these criteria.

Let me give two positive examples of this. The first is a psychologist put jellybeans in a jar and asked individuals to guess the number inside. Each individual answer was significantly off, but when all answers were averaged together, it was surprising how close they came to the truth. The second one comes from Malcom Gladwell's book "Blink". In it he describes how market research and focus groups that are used to predict Top-40 hits were very tough on a singer named Kenna, yet the music industry claimed that his was the most innovative sound they had heard in a long time. Gladwell asks how can focus groups differ so much from experts? The answer is what I wrote above - focus groups are great when you want to appeal to the masses (and I know nothing more mass-oriented than Top 40 Radio), and that's what the research is geared towards.

It is worthwhile to note that what Gladwell fails to ask is whether it was actually a good thing that the groups were so negative towards the singer. If indeed standard groups produce "mass-oriented results", then what the groups are saying is that this person will not be a hit on Top-40. Maybe, however, there are other marketing avenues to get this singer across. Perhaps the masses do not hear what the record executives hear, and therefore Top-40 is too wide an audience. That to me is what the focus groups are saying. It is not a negative that the groups did not like the Kenna - in fact he even says "the problem is that this type of music requires a leap of faith." The fact is though, that Top-40 programers do not take leaps of faith.

So, this brings me to one-on-ones. In these settings, people make decisions independently, and the moderator has more time to probe deeply. Now what do I specifically mean by probing more deeply (focus groups often claim that they do "deep dives" into people's opinions, so how much more deep can a one-on-one interview get?) Well, Gladwell in "Blink" says that when people evaluate new and unique products, they don't have the words to describe them, or relevant frames of reference. As such, people need time to develop these - a moderator can take the time to develop this language with an individual participant more easily than in a group setting. A moderator is able to initiate more challenges and probes in a one-on-one interview.

Evaluating a new product or even a new signer in a group does not leave room for individual opinion. Moreover, in a group, people are not willing to take the "leap of faith" that Kenna recognizes is necessary to enjoy his music. A one-on-one will determine whether there is long-term success for a product. A group will help determine if there is mass, instant-term appeal. The issue for marketers is to determine which is more important to them. Unfortunately, in my opinion as a consumer, the latter seem to dominate.