Wondering how to deal with internet trolls without losing your cool or your community? Don’t feed them: set clear rules, reply once and factually, and ban only as a last resort.
Handling trolls is one part of online community management, so this guide stays practical: what to spot, what to say and when to step in.
| Item | Value |
|---|---|
| Share of communication carried by the words alone | 7% |
| Recommended number of replies to a troll, if you must reply | One, brief and factual |
| Times when the troll rate rises online | Mondays and evenings |
| Place of banning in your response | Last resort |
These figures come from IONOS’s guide to trolling and Loomio’s guide to encouraging good behaviour online. Guidance and findings can change, so check the sources before you rely on them.
What counts as an internet troll?
Global Voices describes a troll as someone who posts controversial, inflammatory, irrelevant or off-topic messages in an online community, with the main aim of provoking an emotional response or disrupting on-topic discussion. You can read that definition in the Global Voices commenting guidelines, last updated March 31, 2019.
Not every negative comment comes from a troll. IONOS points out that an unhappy customer venting is a different case, and that person deserves a real answer rather than a block.
According to the IONOS guide, these signs help you tell the two apart:
- The comment ignores the topic and consists of wild insults, provocations or personal attacks.
- Quotes are twisted or taken out of context, and claims cannot be backed by a reputable source.
- One question is asked again and again, whether or not it relates to the discussion.
- The tone is self-centered and exaggerated, with words like “never” and “everyone”, heavy capitals and swearing.
Timing can matter too. IONOS cites a study by researchers at Stanford and Cornell University which found that people are in a worse mood on Mondays and in the evening, and that the troll rate rises at those times. If you moderate on a schedule, those are the moments to watch more closely.
Prevent trolling before it starts
Loomio’s view is that prevention is the best cure. It points to forums such as MetaFilter and StackOverflow, which combine technical and cultural features to favor respectful participation over trolling, as described in the Loomio guide from 2016.
The same guide names three levers you control:
- Moderation: an engaged community manager who holds space, redirects off-topic posts and enforces boundaries when needed. Without clear boundaries, Loomio says, you exclude everyone who does not want to expose themselves to harassment.
- Invitation: be explicit about what people are being invited into. A vague call to random internet users makes trolling much more likely.
- Identity and commitment: the more committed members are, the less you have to worry about them acting out, so build a shared purpose in open groups.
Written rules turn those levers into something members can see. Social Media Examiner advises posting your etiquette somewhere visible, keeping guidelines reasonable but not too harsh, and telling people what happens when they cross the line, such as deleted comments or a ban, in its guide to social media trolls.
How to respond when a troll shows up
The advice that nearly everyone agrees on is don’t feed them. Global Voices notes that very little can be gained from engaging a troll, because the primary intent of the commenting is to bother you.
When you still need to act, work through these steps in order:
- Pause before replying. Loomio’s advice is not to answer with the same emotional response the troll puts forward.
- Read between the lines. Loomio notes that trolls are often frustrated idealists who may have a legitimate point, so pick out the facts and ignore the heat.
- If you must reply, answer once, briefly and with facts, as IONOS recommends. Global Voices puts it the same way: correct the facts and nothing more.
- Point to your code of conduct and give a friendly first warning, because everyone has a bad day sometimes.
- Own real mistakes. IONOS says a bitter comment can hold a grain of truth, and admitting it publicly takes the wind out of the troll’s sails.
Ignoring does not always work for brands. IONOS warns that ignoring a troll does not mean your followers will, and trolling can escalate once someone steps in to defend you.
Tone is another trap. Loomio notes that we only get 7% of communication from the words themselves, so a short call or in-person chat can settle what looks like hostility in text, per the Loomio guide.
When to block, delete or ban
Some behavior needs removing, not answering. IONOS lists blocking as the first practical tool: it works across social networks, and you can also report comments as inappropriate or offensive.
Deleting works best when you have rules to point to. IONOS warns that deleting content without comment can backfire, so publish a netiquette that explains which comments may be removed and when users may be excluded.
A permanent ban sits at the end of the ladder. Loomio says banning is a last resort that should not be taken lightly and should follow your code of conduct so it is fair and consistent. A sensible escalation looks like this:
- Reply once with facts, or do not reply at all.
- Give a friendly warning that links to your code of conduct.
- Delete the comment and say why.
- Block or report the account.
- Ban, only if the earlier steps have not worked.
Consistency is hard to keep by memory alone, and the right community management tools can help your moderators apply the same rules the same way each time.
Using humor and facts without becoming the troll
Humor can defuse a situation when it is well judged. IONOS calls it a fine line and says it must stay honest and authentic, while Social Media Examiner shows two brand examples in its article dated February 3, 2015.
Sainsbury’s answered a joking criticism about chicken in kind, and the reply drew 141 retweets. O2 disarmed a troll with a humorous reply that was retweeted more than 200 times, according to the same article.
Facts work against rumors. Social Media Examiner argues that the best way to disprove tales told by trolls is with facts. In its example, Apple answered the #bendgate rumors by saying it had received only nine complaints about bending.
Be careful not to cross over yourself. The same piece describes Tesco Mobile’s #nojoke campaign, which poked fun at critics, as getting a lot of engagement but being borderline obnoxious.
Checking whether your approach works
Keep a simple log of what you remove, warn and ban, and the reason for each. A record makes your decisions easier to defend and shows whether the same people or topics keep recurring.
Pair that log with your online community metrics to see whether regular members stay active after a clean-up. If good members go quiet, your rules may be too harsh or too slowly enforced.
Frequently asked questions
Should I ignore trolls completely?
Show answer
Often, yes. IONOS notes that when trolls get no attention they often move on. If your followers are watching, a single brief, factual reply can work better than silence.
How can I tell a troll from an unhappy customer?
Show answer
Look at whether the comment engages with the topic. A critic refers to a real problem, while a troll tends to offer insults, twisted quotes or a repeated question, as listed by IONOS.
Is it fair to ban a troll?
Show answer
It can be, if it comes last. Loomio says a ban should follow your code of conduct so it is fair and consistent.
Can I use humor on a troll?
Show answer
Sometimes. IONOS says that used correctly, humor can silence a troll and show your followers a friendly side, but it is a fine line, so keep it honest.
