Online community numbers

The Center for the Digital Future at the USC Annenberg School recently came out with a new report titled, “Surveying the Digital Future”. This year’s report had an interesting module looking at on-line communities and social networking. The full report is available for sale but the summary has some interesting pieces of data.

Online communities and offline action — The Digital Future Project found that involvement in online communities leads to offline actions. More than one-fifth of online community members (20.3 percent) take actions offline at least once a year that are related to their online community. (An “online community” is defined as a group that shares thoughts or ideas, or works on common projects, through electronic communication only.)

Social activism – Participation in online communities leads to social activism. Almost twothirds of online community members who participate in social causes through the Internet (64.9 percent) say they are involved in causes that were new to them when they began partic ipating on the Internet. And more than 40 percent (43.7 percent) of online community members participate more in social activism since they started participating in online communities.

Online communities: daily use — A significant majority of members of online communities (56.6 percent) log into their community at least once a day.

Member interaction — Online communities are online havens for interaction among members; 70.4 percent of online community members say they sometimes or always interact with other members of their community while logged in.

This is fairly interesting set of data, especially the last bit related to member interactions. The 70.4% number mentioned above seems to contradict the 90-9-1 rule proposed by Jacob Nielsen. I followed up with the authors of the Annenberg study to find out an explanation for the significant difference. I got a prompt response back from Michael Suman of UCLA (Thanks Guys, you rock!!).

Seems that the 90-9-1 rule was developed back in the 90s, which is an
eternity ago in terms of the net. Maybe that has something to do with the


Not an entirely satisfactory explanation…This is still a pretty big discrepancy. I followed up with Jacob Nielsen to get his take on this and got a prompt and detailed response (Thanks Jacob).

Based on their press release, the Annenberg project was only a survey, meaning that it’s based on what people say, not what they do. This means that responses are highly biased in favor of socially desirable statements: people are likely to self-report much more activity and engagement than they actually exhibit.

Also remember that the same person can be a lurker in many contexts and active in a few. If a simplistic survey question asks whether they participate, then the answer could be a truthful “yes”, even if were “no” for 90% of the contexts.

Going beyond the inherent weaknesses of any survey, this survey has been contacting the same people every year for six years. While this provides the benefit of longitudinal data, it also risks increasing the bias, as people may only stay with the survey if they are interested in the kinds of things that the survey asks about.

Finally, what do they mean by “members of online communities”? Quite likely this only includes the 10% of users who are active, and not the 90% who are lurkers, for any given topic. Certainly, the statement that “43.7 percent of online community members participate more in social activism since they started participating in online communities” seems to indicate a very narrow slice of society, since most people don’t participate in social activism.

You highlight the finding that “70.4 percent of online community members say they sometimes or always interact with other members of their community.” First, we don’t know whether they in fact do this, because it’s only self-reported responses to a survey, as opposed to empirical observation. Second, we don’t know how *much* people do this, even when they do. Let’s say that “community members” are the 10% non-lurkers. Then it could easily be the case that 9 of these 10 percent very rarely interact, making the last 1 percent of hyper-active users responsible for most of the interaction.

The fact that participation inequality was named in the 1990s is not a reason to disbelieve it. It’s a reason to believe that it’s a fundamental characteristic of human behavior since it has also been found in systems in the 1980s and 2000s, including the current website examples I mentioned in my article. The older an insight is, the more likely it is to be true, because if it were wrong, there would have been time for lots of studies to accumulate contrary evidence. It’s much more likely that the “latest, hot” press release has some methodological flaw if it contradicts a long-established finding.

Jakob Nielsen, Ph.D.
Nielsen Norman Group

This is getting interesting…What do you guys think?


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s