Here's a few that come to mind:
1) Lurkers are bad, active participants are good. Lurkers offer no benefit to the community. They rarely become active participants. They consume with contributing. They rarely share community content (the people most likely to share are the contributions. The only reason the lurkers are seen as good is because most community professionals don't know how to convert them into active members. It's a defensive argument to pretend bad practice is good practice. This doesn't mean lurkers are necessarily bad – they're just a zero.
2) Big is bad. Bigger communities are worse communities. There is less motivation to contribute. The sense of identity drops. It's tougher to organize collective action. Members join groups that increase their self-concept relative to their peers. If everyone is a member, there is less motivation to join.
3) The community space is not rapidly growing. There is no data to support this and some to suggest it's contracting. We're not shrinking rapidly, but we're not rising to the mainstream popularity as much as we think.
4) Communities don't usually generate a measurable, positive, ROI. We've measured this with a few organisations, the results are mixed at best. We might need a longer time-frame for measurement. For now, the majority of communities aren't generating the extra $200k in sales most need to justify the outlay. This is the result of bad practice, poor measurement, and (often) a bad idea to develop the community in the first place.
5) Gamification doesn't increase participation. Handing members a stack of points and symbols to represent status ensures members only participation for extrinsic reasons. This leads to members with the lowest self-esteem participating the most. It's bad for your members' psychological health too.
6) Personal welcomes are largely a waste of time. They have very little impact upon whether someone becomes a regular or not. Sending a greeting to every member is not a good use of time and has limited impact. Sending messages after a member has made their first contribution has a bigger impact.
7) Forums are probably dying. Most of the data suggests that traditional forums and some of the cheaper community platforms are dying. New entrants, Discourse, Huddler etc…have unique business models that might prevail. The rest are behind and don't seem likely to keep up. No organisation today is launching an online forum to host their community. A forum-like structure will be the bedrock of interactions for many platforms to come, but the notion of the sole-forum platform is of a bygone era.
8) Most communities fail. Look up press release sites for the phrase "launches an online community". You will get thousands of organisations who have launched a community. Now check how many are still alive, most are long-gone.
9) Members don't care about privacy and security. With the exception of mixing identities (i.e. work and personal) and a few topic-specific sectors (e.g. health), members really don't care much about their own privacy or security of their accounts.
10) Communities should not be easy to join. The easier a community is to join the less meaning the act of joining a community has. Most people join and never participation for this very reason. Incentives to join, making the registration form simple to complete, ensuring and automating the speedy process is the best way to get the most people through – but it's not condusive to getting them to participate. You want to weed out the bad people here and ensure the good are well-motivated to be involved.
I'm sure there are plenty more.
You have 45 days remaining to buy your tickets.