← All posts

Trust levels: how the community moderates itself (and why that matters)

Jetonomy's 6 trust levels automatically promote members based on participation quality. How they work, why they cut moderation time, and the design decisions behind them.

The reason most community managers burn out is not “the community is too big to moderate.” It’s “every action requires a human decision, even the obvious ones.” A spam post gets flagged → mod opens flag → reads post → confirms spam → deletes. Repeat 50 times a day.

Jetonomy’s trust level system was designed to remove the human from the obvious-decision path. Members earn moderation abilities automatically based on their contribution history. The community polices itself for the predictable cases. You step in for the genuinely judgmental ones.

The 6 trust levels

LevelNameHow they get thereWhat they can do
0NewDefault on signupRead everything; post in spaces with default join policy
1BasicAfter first few posts + first repliesUse mentions, post links, upload images, edit own posts beyond default window
2RegularSustained participation over timeEdit own post titles; flag content; access trust-2 spaces if any
3MemberSignificant contributions + accepted answersEdit titles of others’ posts (typo fixes); close own duplicate topics
4TrustedLong-term active member with high signalHide spam without mod approval; rename categories suggestions accepted
5LeaderReserved for moderators-by-defaultFull moderation queue access

Promotion is automatic. The defaults are tuned so that genuine spam never reaches trust 2, and a real contributor reaches trust 3 within a few weeks of active participation.

What this changes operationally

Most forum plugins require you to manually grant moderator status to “trusted” members. This is exhausting at scale. You have to decide who’s trusted, remember to demote when someone goes inactive, and field requests like “Can I get mod status to clean up the wiki?”

Trust levels remove all of that. The system grants moderation-adjacent powers gradually as members earn them. A member who’s been posting helpful answers for 3 months can edit a typo in someone’s title without you ever knowing. Another member with 5 days and 2 posts cannot — because they shouldn’t be able to.

The design decisions that matter

Why 6 levels (0-5) instead of 3? Granularity. With 3 levels, the jump from “Basic” to “Trusted” is too big — you have to make a one-time decision about who deserves it. 6 levels means the system can promote in small steps as members earn them.

Why automatic, not manual? Decision fatigue. Every manual promotion is a decision you have to make. Automatic promotion removes that. You can override (demote a misbehaving member) but you don’t have to think about it daily.

Why does trust grow with accepted answers specifically? Because accepted answers are the highest-signal contribution in a community. Anyone can post. Most people can reply. Only members helping solve real problems get their reply marked as the accepted answer — by the person who posted the question. That signal is hard to game.

Why no points / badges as the primary signal? Points and badges are great for recognition (Pro adds custom badges) but they’re easy to gamify. Trust level is calculated from quality signals: accepted answer rate, post survival rate (not flagged/deleted), time on site. Members can’t farm trust by spamming low-quality posts.

What trust levels do NOT do

  • They don’t replace your moderation queue. Pre-trust-2 posts can still flow through Akismet + your flagging system.
  • They don’t replace your judgment for hard cases. A member at trust 4 reporting another member at trust 4 still needs a human to look at it.
  • They don’t automatically resolve disputes. Trust levels grant capabilities, not authority on contested moderation decisions.

The system handles the obvious cases so you can spend your moderation time on the actually hard cases.

How this scales

The math: a community of 5,000 active members typically has ~50 at trust 3+, ~10 at trust 4. Those 60 members handle most of the obvious-case moderation collectively. You handle the 5-10 escalations per week that need human judgment.

This is the difference between “I spend 2 hours/day moderating” and “I spend 2 hours/week moderating.” That’s what shipping trust levels is actually for.

A worth-noting design detail

Trust progression is tied to actions, not arbitrary thresholds. A member posting 100 spam replies will not reach trust 2 because spam replies don’t survive — they get flagged + removed, which counts against the trust calculation. A member posting 10 high-quality answers with several accepted will reach trust 2 fast.

This makes the system resilient to bad-faith actors. They can’t farm their way to mod-adjacent powers. Only genuine contributors do.

How to think about it as a community manager

You’re not “delegating moderation.” You’re letting the system surface the members who’ve already been doing the moderation-adjacent work — answering questions, marking duplicates, helping newcomers — and giving them the tools that match.

This is closer to how real-world communities work. The members who put in the work get the trust. The system just makes that automatic instead of requiring you to track it manually.

Try it

Trust levels are visible in the admin → Jetonomy → Users panel. You can see each member’s current level and what they need to do to reach the next one. Spin up the demo and walk a fake account through the levels to see the system in motion.