Tom Webster, writing and speaking

The Death Of Focus Groups?

Added on by Tom Webster.

Today I watched a little of the coverage for The BIG Conference, a social media event held in my home state of Maine, by eavesdropping on all the folks posting highlights to Twitter. One of the things I caught was a speaker proclaiming that the "Focus Group was dead." Since I didn't actually attend the conference, I won't quote someone I didn't even hear. But it's a sentiment I have heard before from other speakers and writers, so it's the argument and not the arguer I'll touch on here.

The argument against the Focus Group essentially states that better information--more observational, in situ data--is available by mining through the wealth of unstructured data now available in the data streams of various social networks.

There are compelling aspects to this argument--in particular, the belief that "tweeted" comments about brands are somehow more authentic (and certainly less artificial) than similar comments derived from qualitative research constructs. There is an undeniable truth to this--though I would argue that the very act of "Tweeting" to followers carries with it an implied and palpable Heisenberg effect for all who tweet--The Twitterer knows someone is listening, so the Tweet is subtly changed; unconsciously adapted for an audience the author may never know, but wants to please nonetheless.

Consider this, however. As seductive as all that unstructured data is, where do you think it came from?

The 11% of Americans who post status updates?

The 8% of Americans who contribute to blogs?

The 5-8% of Americans who post to Twitter (your guess is as good as mine here, given the number of duplicate/SPAM accounts)?

The 1 in 5 Tweets that are about brands?

The 1 in 5 Tweets about brands that actually express an opinion about those brands (i.e., really 1 in 25 tweets)?

When you consider that there is nearly 100% overlap between all of these groups, you are left with the inescapable conclusion that fewer than 10% of Americans are contributing unstructured data about brands, which leaves the other 90% essentially voiceless in this particular model. My neighbor, for instance, doesn't post to Twitter, only uses Facebook to share family photos and eavesdrop on her kids, and certainly doesn't blog. She is, however, a professor at Duke and drives a nicer car than me. Don't you want to sell her some stuff too? The fact is the vast majority of Americans are online, but don't post about brand experiences online, and going exclusively by the percentage that does share brand opinions might be useful in some ways, but might be horribly misleading for a company seeking to skate where the puck is going. For every PayPal there are 10 Flooz/Beenz-alikes, and calibrating the opinions we can aggregate online still benefits from an offline reality check.

I think the smartest thing one could possibly say about this is that in every focus group I've ever moderated, there are 2-3 vocal, opinionated peer leaders, 5-7 that will go along with the crowd in public, and 3-5 that won't go along but won't challenge the room. As a moderator, I can see and feel this palpably, and get to the heart of the true opinions in the room regardless of the articulation gap that may exist between respondents. If I mine unstructured data, I would only get the former--and worse, I wouldn't really know how many of the latter groups existed. Sample is everything.

So, the non-response bias for "killing" the focus group and other research projects is enormous, and incalculable. Yet, the unstructured data that we can glean from social networks is potentially very valuable, and absolutely supplements and in some cases supplants observations from other forms of market research. Let's not lose sight of that--I'm not--but let's see it as adding incremental insight, and not as the sole source of consumer insight.

I think a better way to think through this is not to use zero-sum thinking. Social networking has enriched my life, and given this quirky introvert a whole new way to express himself before his friends, peers and even potential clients. But it didn't replace the relationships I had before, or how I built and nurtured those relationships. It just made them richer.

Unstructured data makes consumer insight richer--appreciably so--and any market researcher worth their salt will use it. But let's use it to make our focus groups richer, and our surveys more informed--not to exclude the reluctant majority who don't contribute brand opinions online and may not have shared experiences with those who do. Let's use them both, smartly, to create a substrate of data that can provide more actionable and useful market research than ever before. That's what gets my juices flowing.