Soft Techniques in Community Moderation
We’ve seen many community managers work through these issues, and we’ve done it ourselves. There are no set of hard-and-fast rules for community management simply because every community is different. What is acceptable on one community can be unacceptable on another. But generally speaking, we’ve seen some very basic strategies that have been adopted by some of the most successful communities on our platform, and we want to share those with you.
1. Create an explicit set of rules and regulations for your community, publicize them, and strictly enforce them.
You need to define what is and isn’t acceptable behavior on your community so that when the issues inevitably rise, you can point to the rules when taking an action against a community member. These rules typically include guidelines for profanity, image ownership, pornography, and subject matter. Here are two of the most important rules that we believe every community should implement:
- Questioning a moderator decision is grounds for immediate banishment.
- Never engage trolls in any way (this is discussed in detail below).
Here is an example of rules & regulations on the Penny Arcade forum.
2. Promote your good users.
At some point every community manager needs to identify the good and bad contributors so they can promote (ranks) or demote (ban) them, respectively. We break positive contributors into the following categories:
Ambassadors are people who are incredibly devoted to the community, but don’t necessarily contribute the content that is going to be potentially viral. These are people that can be recruited by the community manager to do:
- Content promotion: share great content from the forum on external blogs, social networks, etc.
- Welcoming committee: Greet new members as they join, inform them of rules & regulations, answer general questions about the community.
- Street team: Real-world volunteer duties at conferences & get togethers, street team events, etc.
These are the people who read every single post on the community, watch for people breaking the community rules, and take action to punish those that break the rules. These people need to be unemotional about their implementation of the rules, never engage in discussion with people being punished, and simply enforce the rules as they are clearly laid out.
These are the viral content creators who provide unparalleled expertise in the subject matter at hand. They know more about the subject matter of the community than the people who run it. Their content should be automatically highlighted above the content of others, they should have special abilities (ranks) to post different kinds of content or have special functionality above others.
Use ranks to promote Super Users and Ambassadors. Ranks give users the ability to post different kinds of content, have their content highlighted above others and various other non-administrative abilities. Put enforcers into the moderator role so they can take real administrative action against others when they break the rules. Obviously make sure you trust your moderators to not break the rules themselves, enforce the rules to the letter, and not take advantage of their permissions to promote their own content.
3. Weed out the bad users.
We break negative contributors down into the following categories:
There are various types of spammers. Some post legitimate content only to come back and change it after it’s gained some SEO juice. Others outright spam content wherever they can. Others include spam in their signatures.
Hackers can often be very difficult to identify simply because by the time you know they are present, it is because they’ve already successfully hacked their way into someone else’s account and are using it to post negative content.
Some trolls know they are trolling, others don’t. Both are community poison. The golden rule when dealing with a Troll is to not engage them in any way because their goal is to get any response (positive, or preferably negative) out of people.
Spammers and hackers should be banned without exception. Trolls are trickier. Because trolls want to illicit a response from people, responding, warning, and banning are all ineffective against trolls. They will happily take a tongue lashing and banishment only to re-register and start the process all over again. In Vanilla we use reactions to identify trolls. Once identified, a system account automatically bans the trolls. Since trolls want to get a response from a human, being banned by software implementing business rules is incredibly effective.