Enter your email address:

Delivered by FeedBurner

About This Blog

Josh’s Tweet Stream

  • More tweets

« Want to give your clients signed copies of Groundswell for the holidays? | Main | What Will Be In GROUNDSWELL HEROES »

November 30, 2009

Research in a petri dish: Learning from communities

by Josh Bernoff

From my Marketing News column.

The problem with market research is that, while it can generate some pretty actionable insights, it’s expensive. If you do surveys, it’s always a struggle to get enough people to generate a large enough sample without breaking the bank – and when the survey’s done, pray you don’t find out that there was some question you forgot to ask. For qualitative feedback, there’s always the trusty focus group, but to make sure those insights are on target you’d better do eight of them all around the country, and there goes the budget again.

There must be some way to do research with a group of your customers who want to stay in touch, are on call whenever you want them, and respond well to follow up questions.

In fact, there is. It’s called community research. Here’s how it works. You recruit 300 people, or 1,000 people, into an online space. Like a focus group, they know they’re there because they fit your description, and to help you make decisions. But unlike the focus group, it’s an ongoing relationship. So you can ask them for help continuously.

If this sounds like a lot of work, it is. While you can do it yourself, it’s a whole lot easier to work with a vendor that runs these research communities, like Communispace or Passenger. They’ll recruit the people, moderate the community, and help you generate the insights. Expect a six-figure annual bill. But what are you spending now for surveys and focus groups? Read on, and then figure out if this is a better, possibly cheaper substitute.

Here are some real-life cases of companies that have used research like this.

Hyundai convened a community of car owners (built and run by Passenger) and asked them about programs to help buyers affected by a recession. “Should we let you return the car if you lose your job?” they asked. “Actually,” the community replied, “If we lose our jobs, we’re going to need the car to find a new one.” Result: the Hyundai Assurance Plus program, which actually makes up to three car payments if you lose your job.

Intercontinental Hotel Group has three research communities managed by Communispace, and they’re a key part of the company’s market research – and marketing. The frequent travelers in the group contributed real-life photos – not only did IHG save on stock photo fees, but the collateral has generated a 24% lift in room revenue. IHG also learned that community members had no interest in Twitter as a channel to communicate with the company, so the hotel chain scrapped that part of their social media plans.

Companies are even using communities for product design. Charles Schwab used a Communispace community to decide that its checking account should include fee-free ATM use. Del Monte Pet Products used a MarketTools community to design on the shape and ingredients of a doggie snack, Snausages Breakfast Bites. TurboTax’s popular bookmarks feature came from a discussion that members initiated themselves in the company’s “Inner Circle” tax prep community.

Naturally, research communities are not a panacea. With only a few hundred members, they not as representative as demographically balanced surveys. You’re hearing from the loudest members of your audience; shy people will be poorly represented. And some of the emotional elements you can get from a focus group may not come through as clearly if you’re not a pane of glass away.

But research communities can do things no other form of research can do. You can ask followup questions to get clarification or more detail. You can look at the profiles of members, to put their questions in context. And most interestingly, you can see what they want to talk about. Although there are plenty of moderator-initiated questions in a research community, most of the chatter is member-generated, and much of the most interesting learning comes from those discussions.

If you’re going to go forward with a research community, there are a few questions you should ask yourself (and if you want more detail on this, see Forrester’s document on the topic, “Use Online Communities For Strategic Insight” by Lisa Bradner and Cynthia Pflaum. [Full access for clients.])

First, decide if you want to focus on traditional research or innovation. Communities for traditional research should match your target group and complement your other research. Innovation communities should be stocked with more vocal types.

Second, decide if the community should be private or public. Private communities, like the ones Communispace runs, aren’t visible on the open Internet, which keeps those insights away from competitors. But public communities can attract a much larger audience and grow organically.

Third, figure out exactly whose going to be in your community. A community of customers can help you with issues of loyalty and usability, while a demographic community (say, moms) can help you see what marketing will work best with prospects.

And if you worry about whether these collections of consumers function like real communities on the Internet, you shouldn’t. One of the most amazing stories we’ve heard was from the research community of cancer patients run by Communispace for the National Comprehensive Cancer Network, a group of cancer centers including Memorial Sloan-Kettering. Having learned enough from a group of early-stage cancer patients, the cancer centers started a second community of more advanced patients, and asked Communispace to shut down the original. But the members of the original community objected – they didn’t want to leave their friends behind. To them the community wasn’t a research project, it was a support group. (Communispace and NCCN left it running, of course.)

Do your other research tools generate passion like that? Not likely.

TrackBack

TrackBack URL for this entry:
http://www.typepad.com/services/trackback/6a00d8341c50bf53ef012875f48b86970c

Listed below are links to weblogs that reference Research in a petri dish: Learning from communities:

Comments

Julie Wittes Schlack

Thanks for so practically articulating the benefits of online research communities, Josh. I do want to offer a different perspective on one small point, though, which is your assertion that in communities you hear only from the loudest people, and emotional content can be absent. While I think that hearing only from the "loudest" is a risk of public and non-facilitated communities, in the private, highly facilitated communities that we at Communispace run, we're very mindful and systematic about drawing out everyone -- through our daily moderation, by enabling members to post anonymously if they're more comfortable doing so, and by providing multiple modes of self-expression. And we intentionally exclude ratings, thumbs up/down icons, and other standard elements of many social networking sites that could have a potentially chilling effect on honest and frequent participation. As for emotional expression, the intimacy of our relatively small, private communities coupled with our abilities to let members express themselves through text, images, video -- even voice mail -- generates really rich emotional nuance. Our members tell us that unlike public forums, our communities feel "safe."

Ali McCourt

Hi Josh - Ali from the TurboTax Inner Circle here. Thanks for including Intuit in your examples of how online communities help drive innovation and customer satisfaction. Our Inner Circle is now up to 25,000 members which has helped us address the representation issue. We now have a pretty diverse group of customers who collaborate with us. Our bookmarks feature rolled out to our Mac customers last year and was such a delighter that we've included it in all our TurboTax products this year. Just one small example of how listening to customers can better the product experience!

Kathryn Korostoff

Yes, communities have their place. I ardently believe that there are many organizations that can benefit from structured online communities (MROCs, as many now call them), or even just well-run customer advisory boards (http://bit.ly/5QHHeC).

But, please, some reality checks:

“And if you worry about whether these collections of consumers function like real communities on the Internet, you shouldn’t.” Well, I think what you mean is that “some of them do.” But nobody thinking about funding (and we are talking about significant funding here) a community should assume they are guaranteed to have the same experience. Few communities have the defining bond that cancer patients have. Some organizations simply don’t have customer bases so eager to connect with peers for experience sharing, support, etc. Some do—there are some brands that have very passionate customers (Apple comes to mind). So be real: does your organization have a customer base that will be self-motivated to participate? If so, great!

The statement, “But research communities can do things no other form of research can do. You can ask follow-up questions to get clarification or more detail. You can look at the profiles of members, to put their questions in context”, is factually incorrect. You can do that with most types of research. Any half-decent researcher knows to gather contextual information (either for focus groups, interviews or survey projects), and in many cases the screening process uses techniques to ensure this context is valid. Doing online research? There are most certainly tools available that allow you to do real-time probing (check out Invoke Solutions and iModerate—two very cool companies). Yes, online communities can do these things too—but to say “no other form of research can” is blatantly untrue.

MROCs have a place. But whenever we gather customer insights, or insights from the broader market, we always need to first ask, “What are the objectives? What do we want to learn? What will we do with this data?” The answers to those questions will drive the key decision about best methodology. Should the research be “blind” (the sponsor kept anonymous to avoid bias)? Should it be quantitative? Is directional insight sufficient? Should it be done in one country or many? Should it include our customers and/or competitors’ customers? Only then you can pick which tools, or combination, will be the best choice for your market research dollars.

BTW, I loved Julie’s comment: “And we intentionally exclude ratings, thumbs up/down icons, and other standard elements of many social networking sites that could have a potentially chilling effect on honest and frequent participation.” This is so important.

I welcome further discussion via email or my blog http://www.researchrockstar.com/blog/

Josh Bernoff

Thanks for the very thoughtful comments, Kathryn!

In working with providers like Communispace, many clients tell us they've found that they can get people engaged, even if the product is boring. That is, moms may not want to talk about toothpaste, but they may want to talk about keeping their kids healthy. If you don't design the community to engage, you'll suffer lack of participation, just as you say.

Also, I don't know of any form of research that allows people to come back a week later and ask follow up questions. Surveys don't do it. Focus groups don't do it. You'd need a standing panel of some kind -- and that's just what an MROC is.

Absolutely, these MROCs are not for everybody. And they can be quite costly. But for many projects, there is no other way to get what you're looking for, if it's deeper ongoing insight you want.

michael  louca

They are useful tools as long as they don’t replace traditional quant market; and should are used in the context of a good quant research program---like Technographics ☺---that can put findings in the right context.

I also wonder if the way a lot of these “communities” are created and managed get you too focused on customers you already have and not on customers that you are not reaching.

Kathryn Korostoff

Thanks Josh--and nice to meet you. Regarding your comment: "Also, I don't know of any form of research that allows people to come back a week later and ask follow up questions. "
I've done that with survey projects many times. Sometimes you post a follow-up survey to the respondents of the original survey, or simply conduct follow-up phone interviews with a subset of the respondents. And if you did online focus groups or OLBBs, you can certainly do it easily (unless you recruited from a list provider that prohibits that--which is sometimes true). Traditional focus groups: not the case, since they are usually recruited from facility databases.

Stephen Cribbett

Hi Josh, Great post. Could you shed some more light on how members of these online communities are rewarded and incentivised for their participation, since traditional research approach to these might prove very costly.
Thanks.

Josh Bernoff

@Stephen Cribbett Some times members of these communities are not compensated at all. In other cases they get a small stipend (like a gift certificate) a few times a year. The main cost is in managing the community, not compensating the members.

Jen Stevens

Hi Josh,

Great post and very insightful comments from your readers as well.

You point out the dilemma -- surveys on one end, high end focus groups on the other. Both ineffective in their own way. Given the fact that you have to leverage the community of users/consumer for feedback and insights, how to you get what you need?

Regardless of the branding and labeling of the methodology, our customers have achieved great success by making the following a part of their strategy -

1. Acknowledge and accept that you don't control the users/consumers and cannot make any meaningful demands of their time. You earn their respect and value any and all insights they provide.

2. Convert every interaction into a user centric conversation - the key words are *user centric* and *conversation*. Its really not about your product or service, its about how the consumer felt about the experience. And, conversation involves *listening* -- probably much more than speaking or broadcasting.

3. Don't *limit* the feedback (your points about why invitation based communities are not a panacea), but rather analyze the insights based on the who, what and more importantly why.

We at IdeaMagnet are glad to provide the technology platform to power these initiatives. We are early in the game but are encouraged by what our customer have been able to accomplish.

Its great to see this post reaffirm some of our findings and observations over the past few quarters. More information about IdeaMagnet is available for those interested at http://idea-magnet.com/conversations_not_surveys

Thanks,

Jen Stevens

Типография

I also wonder if the way a lot of these “communities” are created and managed get you too focused on customers you already have and not on customers that you are not reaching.


The comments to this entry are closed.