The practices of online community management: the results are in

Online communities are all around us. I myself spend a lot of time on Edgeryders and Spaghetti Open Data, and there are countless others. Most of these communities are managed in one form or other – in fact, both online communities and their management predate the Internet itself. But what do online community manager actually do? How do they spend their time? There is quite a lot of lore and anecdotes out there, but it is hard to have an idea of how representative they are.

As part of a larger work (my Ph.D. thesis), I decided to try to answer this in a more systematic way. Here is what I did.

Methodology and questionnaire

I started by scanning the academic and business literature on online community management looking for practical advice. From it, I extracted a list of practices. The list is this:

  1. Invite users to join [Young 2013, Iriberri 2009, Kraut 2011].
  2. Welcome new users when they sign up [Kraut 2011, Kim 2000, Ganley 2009, own experience].
  3. Engage with users, to encourage them to be more active and make them feel welcome [Young 2013, Ludford 2004, Williams 2000, Kim 2000, Kraut 2011, own experience].
  4. Mediate conflict (Kim 2000, Kraut 2011].
  5. Tweak and optimize the user experience (if you consider this to be part of community management) [Kraut 2011, Kim 2000].
  6. Encourage interaction between members (Ganley 2009, Williams 2000, Kraut 2011, own experience).
  7. Organize real-world meetings (Kim 2000, own experience).
  8. Acknowledge members for their contribution (Kraut 2011, Ludford 2004).
  9. Support volunteers (Kim 2000, Young 2013, Williams 2000, Kraut 2011).

Then, I prepared a questionnaire that had one question for each practice of the list. The academic references (given in full at the end of this post) were not included in the questionnaire, to avoid influencing respondents. For each practice in the list, respondents were asked to answer (by multiple choice) the following question:

To manage your online community, which of these courses of actions do you take, and how often?

Next, I included a question to allow respondents to point to practices not in the list. Answers were given in free-form text.

Do you want to add any other activity that uses up significant chunks of your community management time?

Finally, I added two more questions for context. Answers were given by multiple choice.

How old is the community you manage? If you manage more than one, refer to the oldest.

How large is the community you manage? If you manage more than one, refer to the largest.

I created a post on this blog that contained some context information and linked to the questionnaire itself on Google Forms. Post and questionnaire went live on March 13th 2018. I created a shortlink via pointing to the information page, and disseminated via my own Twitter and Facebook accounts. I also posted it on e-mint, a long-running Yahoo! group populated by professional online community managers, and on CMX Hub, a community of community managers on Facebook.

I have collected results on March 31st 2018. At that date, my shortlink had collected 210 clicks. reports that 101 came from Facebook; 78 from e-mail or direct; 25 from Twitter. Geographically, most visitors came from the United States (82), Italy (49), the UK (23). These visits resulted in 83 completed questionnaires. This is an amazing result! I am very grateful to everyone who responded and spread the interest for my little initiative. In particular, I suspect that the benevolence of my friend John Coate and the e-minters played a large part in this success.

Results and data

Inviting people to join is practiced by almost all respondents. 68 out of 83 have answered “often” or “occasionally”.

Once users sign up, most respondents send out a welcome message. 47 do it “often”, and a further 9 do it “occasionally”.

Engagement with users is overwhelmingly practiced. Only 5 out the 83 respondents do it “rarely” or “never”.

Conflict mediation is also an important activity. 33 respondents report practicing it “often”, and an additional 22 “occasionally”. Only 28 of them report peaceful communities, where conflict mediation is practiced only rarely or never by community managers. This was a surprise for me. I guess I gravitate towards in unusually peaceful online hangouts!

Almost every respondent engages in design or co-design of interfaces. Clearly community managers think doing so is part of their role.

Almost every one of them is also preoccupied with getting people to talk more to each other.

About nine respondents out of ten allocate at least some time to organizing offline community events, but only 12 of them reported doing it “often”.

Thanking and acknowledging active members is perhaps the most widespread activity, with over 90% of respondents engaging in it “often” or “occasionally”.

Finally, about half of the informants support volunteers “often” or “occasionally”. About a third never does it.

Most of the informants referred to rather established communities with over five years of history. Only 8 of them reported managing new community, created less than a year before taking the questionnaire.

The respondents were quite well distributed by size of the communities they manage. Interestingly, the relative majority manage large ones, with over 100,000 accounts each.

You can download the dataset from The data are open, published under a Creative Commons Attribution 4.0 license.  This means you are welcome to use them for your own research, as long as you cite me as the dataset author.

Full references for the list of practices

Ganley, Dale, and Cliff Lampe. “The ties that bind: Social network principles in online communities.” Decision Support Systems 47.3 (2009): 266-274.

Iriberri, Alicia, and Gondy Leroy. “A life-cycle perspective on online community success.” ACM Computing Surveys (CSUR) 41.2 (2009): 11.

Kim, Amy Jo. Community building on the web: Secret strategies for successful online communities. Peachpit Press, 2000.

Kraut, Robert E., et al. Building successful online communities: Evidence-based social design. Mit Press, 2012.

Ludford, Pamela J., et al. “Think different: increasing online community participation using uniqueness and group dissimilarity.” Proceedings of the SIGCHI conference on Human factors in computing systems. ACM, 2004.

Panzarasa, Pietro, Tore Opsahl, and Kathleen M. Carley. “Patterns and dynamics of users’ behavior and interaction: Network analysis of an online community.” Journal of the Association for Information Science and Technology 60.5 (2009): 911-932.

Williams, Ruth L., and Joseph Cothrel. “Four smart ways to run online communities.” MIT Sloan Management Review 41.4 (2000): 81.

Young, Colleen. “Community management that works: how to build and sustain a thriving online health community.” Journal of medical Internet research 15.6 (2013).

Mapping online community management practices: can I have a little help with my thesis?

I know: I am quite old to be a student (I am 52, if you like numbers). Still, learning is a lifetime thing, and in this period of my life I am doing part of my learning as a Ph.D. student at University of Alicante.

My thesis concerns online communities. I view them in a network perspective; I model them as networks of interactions, and study the network topology in search of the signs of collective intelligence. My real interest, however, are societies at large: I view online communities as “toy societies”, complex enough to be interesting but simple enough for rigorous empirical analysis. In the end, I would like to be part of the adventure of modelling social dynamics. I think that is important, because, if we figure it out, we can use the knowledge to build better online communities, and, in the end, better societies.

It is a fantastic journey: along the way, I learned Python, NetLogo, some R, and quite a lot of network math. I also brushed up on my statistics and econometrics. I also took brilliant courses and wrote neat stuff. But now it is time for this journey to end, and for me to move to the next phase. The good news is that my thesis is almost finished. The bad news is that it is not quite finished. For that, I need your help.

I have been reading up on the literature on online community management, in search of what I am calling policies. These are simple rules of the type “to obtain x, do y“. Which policies does the literature on online communities recommend? I ended up with the following list:

  1. Invite users to join.
  2. Welcome new users when they sign up.
  3. Engage with users, to encourage them to be more active and make them feel welcome.
  4. Mediate conflict.
  5. Tweak and optimize the user experience (if you consider this to be part of community management).
  6. Encourage interaction between members.
  7. Organize real-world meetings.
  8. Acknowledge members for their contribution.
  9. Support volunteers.
  10. Other?

So, here’s how you can help.

  • If you are yourself stewarding an online community, no matter how large or small, please fill in the survey below.
  • If you know someone that does, please send him/her my way.
  • Reshare this link on your Facebook/Twitter: (redirects back to this page).

You will earn the gratitude of an aging student that really wants to grow up to be a scholar!

Salzburg Global Seminar: a social network of session 593

Folks at the Salzburg Global Seminar were kind enough to show interest in (or at least tolerate) my obsession for social networks and semantic social networks. So, I made a social network of our session, called “session 593” (a nice prime number, as Martin Bohle pointed out).

It works like this. There are five types of nodes: fellows (brown), staff (yellow), plenary panels (green), focus groups (blue) and impromptu breakout sessions (red). Staff and fellows “vote” participating in focus groups and breakout sessions. Additionally, SGS assigned many of us to plenary panels with others. Edges in the network are interpreted as “fellow X participated to event Y”.

The data are wildly incomplete. I compiled the lists of fellows, staff, and plenary panels from the program; the list of focus groups I made on the fly on the last day. The program also has data about who participated in which panel, so that’s there. Kiley’s latest two recaps count as panels, because she involved others in them (Katindi, Brenda, Zhouying…). As for the focus group compositions, I obviously knew the one I participated in, thought to action; I also was able to add two more (being human and global lab), based on the tables on the final session. I had started to map the arts and creative practice , but then the facilitator asked us to stand up and move the table, and there went my data integrity 🙂 I also do not know who participated into which session, except for a few (Martin’s, Eichi’s, my own…).

If the data were complete, you could start looking through which sessions connected who, which people spent lots of time together (this is done through a technique called projection), and even, with some reflection, who should have spent time together but did not – the missing edges in the network. With the incomplete data, it turns out that the global lab focus group had the highest eigenvector centrality (a measure of centrality that reflects the centrality of connecting nodes, like Google’s PageRank algorithm). It is also the session with the most participants.

If you were at SGS session 593, and are curious as to what this might look like, I am happy to try to complete it. I also vow to beautify it a bit – makes for a cool pic to put on your blog. I will need:

  1. From everyone, which focus group they participated to.
  2. From people who held breakout sessions, who came to their session.

I will update it as I receive data from you. I predict that the complete data would see a high centrality of Claire (Nelson) and her Moonshot session. 🙂

This network has no semantics, it’s just a social network. But still, networks speak to many people, myself included, and anyway doing something like this is easy.

If you are in the network, and prefer not to be included in the network, let me know and I will remove you at once.