What is the optimal size for a market research online community (MROC)?

I have recently been asked this question with some frequency. This, in itself, represents the level of interest in MROCs and also a certain level of bias from survey-based quantitative researchers.

One avenue of approach on this topic is Dunbar’s number. Dunbar’s number is based on Robin Dunbar’s primate research. The number itself is a mean group size of 148, but with a range of 100 to 230. This represents the “cognitive limit to the number of individuals with whom any one person can maintain stable relationships.”

What does this mean for the “optimal” MROC size?

Dunbar’s number suggests that MROCs in the 150-230 range are optimal and that larger population MROCs may be difficult for participants to develop much of a sense of group identity with.

But, there is a diversity of opinion on this topic. Some, like Jacob Morgan argue that applying Dunbar’s number to MROCs is inappropriate since we’re not trying to build a tribe or fighting unit and are only attempting to get a critical mass for consumer co-creation. The argument here is essentially one asking why we should limit the power of networks.

Some clients and prospective clients feel that limiting communities to a range around Dunbar’s number is too small and that the size of the community should more closely resemble the sample size of traditional national surveys – 800 or 1000 (or more). In this instance, this may be an example of researchers taking one model (projectable surveys) and applying them to a very different tool (directional, qualitative community insights). But, I must admit that community sizes larger than Dunbar’s number are comforting to me as a researcher.

For a deeper dive into this topic, I recommend Matt Foley’s blog on the topic.

Jeff Henning’s blog on this top is also worth a read, as is Jacob Morgan’s.

To me a trade off seems to exist. If it is very important to the research that a tight knit community develop, then Dunbar’s number seems to be a guide. But, if a digital tribe is not critically important, there seems to be a case for a larger network.

I suspect that quant researchers will want a larger community and qual researchers will want a smaller community.