Module 6: Moderation
Moderation is where community management stops being hospitality and becomes judgment. Members are paying customers, so every enforcement decision is also a customer-relations decision. This module gives you the ladder, the conflict playbook, and the safety setup that keep a community firm, fair, and calm — including at 2 a.m. when you're not watching.
The moderation ladder
Amateur moderation has two settings: ignore it, or ban them. Both fail. Ignoring teaches the group that the guidelines are decorative; banning a paying member over a first heated comment teaches everyone else to be afraid of you. Professional moderation is a ladder — small, documented steps that escalate only as far as the behavior requires.
Run it like this. Step one: the quiet nudge. In the thread, light and brief: "Let's keep this on the offer itself, not each other." Most decent people recalibrate instantly when they notice a human is watching. Step two: the private DM. Name the specific guideline, quote their words back, state what needs to change, and leave the door open: "You're a valued member here; this part has to stop." Step three: a formal warning with consequences stated plainly — next step is a posting pause or removal. Step four: restriction or removal, executed without a public scene. Repeat offenders climb faster; someone whose second incident is worse than the first doesn't get to restart at a nudge.
Document every step, every time, even the nudges. A simple log — date, member, what happened, what you did, links or screenshots — takes two minutes and saves you in every hard case. When a removed member disputes the decision, complains to the owner, or demands a refund "out of nowhere," the log is the difference between a calm explanation and an argument about memories. It also protects members: patterns become visible in a log that no single incident reveals, like one person quietly starting every fight.
Two more rules. Enforce against behavior, never identity or opinion — "this comment breaks rule two" ages better than any judgment of character. And apply the ladder identically to the newest member and the loudest superuser, because the group is watching both cases. Fairness you can show is worth more than friendliness you can only claim.
In a feedback thread, a longtime member posts a sarcastic reply that mocks a newcomer's business — personal, but not profanity or threats. It's his first incident. Where do you start?
The conflict playbook
Conflict in a paid community has an audience of customers, so how you handle it is part of the product. The playbook has three moves: cool it in public, resolve it in private, brief the owner fast. In that order, every time.
Cool it in public. When a thread turns hot, your public job is temperature, not verdicts. Slow it down: "Pausing this thread for an hour so everyone can cool off — locking comments, not deleting anything." On Circle and Skool you can limit or close a thread; on Discord you can apply slowmode or lock the channel. Never litigate who's right in the thread itself — public rulings create winners, losers, and a second argument about the ruling. And never delete the evidence in anger; remove only what breaks safety rules, and note what you removed and why in the log.
Resolve it in private. DM each party separately. Let them vent to you instead of the group — half of conflict resolution is being the cheaper place to say the angry thing. Establish what actually happened (screenshots over summaries), name the guideline involved, agree the specific change, and close the loop with both sides: "Here's what I've asked for, here's what happens next." If money is involved — a member accusing another of a bad deal done inside the community — your role is containment and referral to the owner, not arbitration of their business dispute.
Brief the owner fast. Owners hate surprises more than problems. The moment a conflict involves a threat, a refund demand, a public accusation about the business itself, or anything that could leave the platform, the owner gets a short brief: what happened, what you've done, what you recommend, what you need from them. Five lines, same day. An owner who hears about a blow-up from a screenshot in their DMs will never fully trust your coverage again; an owner who always hears it from you first will defend your calls in public — which is the backing that makes the whole ladder work.
Two members are escalating in a thread — one now accuses the other of scamming him in a deal they made through the community, and screenshots are being posted. What's your sequence?
Spam, raids, and safety basics
The cheapest moderation is the incident that never reaches the members. Most of that is setup: settings and habits you put in place once, before the bad night, not during it.
Spam first. Paid communities have a wall — payment — that keeps most spam out, but member accounts get compromised and free-trial or free-tier spaces get hit. Use the platform's defenses: on Discord, set verification levels so brand-new accounts wait before posting, keep the moderation bots from Module 3 (MEE6, Carl-bot) configured with link and flood filters, and restrict link posting for new members where the platform allows. On Circle and Skool, lean on join questions, manual approval for suspicious signups, and keyword alerts on the classic spam vocabulary. Give members exactly one obvious place to report: a report button habit, a mods tag, a named channel. Reports you never hear about become exits instead.
Raids — coordinated floods of disruptive accounts — are mostly a Discord and open-community problem, but any community linked in the wrong place can get one. Prepare the boring kit in advance: know where the panic controls are (pause invites, raise verification, lock channels), keep them in a written runbook you can execute half-asleep, and have a second mod who also knows it. During a raid, protect first and investigate second: lock down, clear the flood, brief the owner, and only then sort genuine members from the wave. Restoring calm visibly and quickly is also retention — members judge the community by its worst night.
Safety basics are non-negotiable and mostly unglamorous: no member's personal information shared without consent, ever; harassment and threats skip rungs of the ladder and go straight to removal plus an owner brief; anything involving minors or credible real-world harm gets escalated to the owner and the platform immediately, not handled as a culture matter. Write these into your guidelines in plain language so enforcement is a reference, not an invention. The communities that sail through bad nights are the ones whose manager did the quiet setup months earlier — this lesson is that setup.
Your Discord community is linked in a big public post at 11 p.m. and dozens of brand-new accounts start flooding the main channel with links. You're the only mod online. First moves?
What you can now do
- Work a moderation ladder — small, documented steps that escalate only as far as the behavior requires — instead of the amateur settings of ignore it or ban them.
- Run conflict by the playbook, in order, every time: cool it in public, resolve it in private, brief the owner fast.
- Set up the spam, raid, and safety basics in advance, so the cheapest moderation — the incident that never reaches the members — is already working at 2 a.m.