OPINION3 November 2011

Failing communities

Trends

“We have made more mistakes than anyone else,” said Communispace CEO Diane Hessan, kicking off her talk at Festival of NewMR 2011 about what can go wrong within market research online communities.

“We have made more mistakes than anyone else,” said Communispace CEO Diane Hessan, kicking off her talk at Festival of NewMR 2011 about what can go wrong within market research online communities.

Here are her five things that haven’t worked over the years:

  1. The bigger is better syndrome: Community size is the wrong metric to focus on. “The perfect size for us, having done about 600 of these communities, is about 300 to 500 members,” Hessan said. “It sounds small, but the metric that matters is engagement. If the community is too big participation actually goes down.” Communispace has a 64% rule for participation rates, which works out at an average of 7.3 contributions per active member each month. Compare that to Facebook Fan Pages, Hessan said. Recent research suggests that for every 100,000 fans a page has only an average of 54 people will actually ‘like’ a post and only nine will comment. “Engagement trumps sample size.”
  2. Survey-itis: Surveying too much can be an easy mistake to make and a major problem. “Over the years we have had communities where we have just surveyed people but members told us this was just downright boring,” said Hessan. “As you increase the variety of activities you increase participation.” A study of 57 Communispace communities found that those with more balanced facilitation generated a much higher number of average monthly contributions. “Until 2009 the preponderance of what we were doing was just discussion boards and surveys,” said Hessan. “But there are so many examples of what you can do above and beyond traditional market research. We have had great success with mobile ethnography, heat-mapping, collaging, whiteboarding, video chats, co-creation. The net here is that by using less researchy, more humanist approaches, success has gone up.”
  3. The garbage pail: Hessan’s term for doing things with a community that are not worthwhile, or succumbing to the temptation to use it for everything – like working out 2+2 on a calculator because the calculator happens to be to hand.
  4. The Trojan horse: sometimes people want to use a community to listen to consumers, but others want to create advocates, Hessan said. “But consumers know if you are only pretending to listen to them, just so they will go out and buzz about you… that has never, ever worked for us.”
  5. The traditional research mindset: Concepts like the importance of anonymity, of keeping things neutral and structured and ensuring the client is invisible to consumers don’t apply in communities, says Hessan. “What works is being open, transparent and having a conversation.”

Brian Tarran

Trends

@RESEARCH LIVE

4 Comments

Charles Pearson - Co-Founder & Managing Director,

14 years ago

Some good advice but the reality is that most consumers have limited time and a larger panel comprised of that broad spectrum and more complete distribution of consumers who will want to participate once or twice per month is a far better choice for most companies. In this way, the results will be more representative and the number of completed interviews for the surveys that need to be completed for the business to address the core questions and decision making will be large enough to be accurate and reliable. All too often these very small panels referred to in this article are being used for quant surveys because that is at the core of the client's needs and they are being completed with 5,10, maybe 25 completed interviews and then being interpreted as if these results were significant.

Like Report

Derek Sawchuk - Passenger

14 years ago

I have to agree with you, Charles, that in most cases size really does matter. I certainly don’t want to blur the lines between community and panel, but slightly larger communities allow for a greater diversity of engagement opportunities. Also, they’re better suited to support dynamic segmentation. As a result, clients can go beyond a standard insight community and achieve multiple research and innovation objectives, and ultimately gain more value.

Like Report

Jay

14 years ago

If you're only using them for quant surveys, what on earth makes it a "community"? It's just a convenient panel for recruitment. Communities get to know each other and interact - that's the very definition of the word. Hardly any market research companies are running communities in this way, but if you are you're certainly using forums and chats and newer qual-er methods - replicating the things people are doing in social media naturally, of their own accord. (Which is to say, no surveys beyond a one-question poll...)

Like Report

Charles Pearson - Founder & MD, EasyInsites

13 years ago

Jay, I disagree. Even large panels have aspects which make them a community. Ask any member of a custom panel (at EasyInsites, some examples include Diageo's Bar Talk, Bounty's Word of Mum, WHSmith's Surveys, and Premier Chefs), and they will tell you that they belong to a community. They have the opportunity to interact by participating in quick polls where they can immediately see how their answer compares with others in the community. They also have the opportunity to participate in various online qualitative research, such as online focus groups and bulletin boards, where they can interact with others. They are also provided with various content, such as news and results, which are exclusively provided to members of that community. I think you need to remember that market researchers are not responsible for brand communities, unless these communities have been built for the purpose of gathering feedback and insights. And why would you want to invest in trying to replicate what the naturally forming communities are doing in social networks when you can just observe these interactions. Clearly, if companies want to watch their customers "get to know each other and interact", the naturally forming ones will be the best way in which to do that i.e., far more real and far more cost effective!

Like Report