How to Report on a Discord Community to Your Leadership Team: The Six Line Weekly Report

Read this if you check the server daily

The habit looks like diligence. It is closer to a missing instrument.

Somebody has to answer the question of whether the community is working. If no page answers it, a person has to, and the founder is the one who volunteers because the founder is the one carrying the risk. The cost is not the fifteen minutes. It is that the question is never closed, and an open question about something you own is the thing that makes you avoid the app.

The replacement is small enough to build this week. One page, six lines, three columns behind every line, sent on the same weekday until it is boring.

What reading the room actually measures

Be honest about the instrument you are currently using. Opening the server measures the loudest channel, at whatever hour you happened to look, weighted toward the last twenty minutes of conversation. It has no memory, so it cannot compare this month to last. And it is blind to the two groups that matter most: the members who arrived and quietly left, and the members who asked something and got nothing back.

Those two absences are where the money is. Neither of them makes noise, so neither of them survives a reading based check.

The six lines, written out

1. Finished entry. Joins this week, next to the number who completed entry and posted once. Two figures, never one. A rising join count with flat completion is a leak with a budget pointed at it.
2. First answer time. The middle case and the worst case for a first human reply. The middle case describes a typical member. The worst case describes the member who goes and tells someone about it.
3. Unanswered past a day. Not a number, a short list of links. An unanswered question is the most expensive object in a community, because everyone who reads it learns what asking gets you.
4. Members who spoke. Distinct members who posted or reacted. This retires total membership, which stops meaning anything after the first month.
5. Moderation load. Actions, escalations, repeat cases. Read it as a staffing and clarity signal rather than a safety signal, because rising load with flat membership usually means a rule is unclear or the gate is admitting the wrong people.
6. Decision caused. One business decision traceable to member input, named. One empty week is fine. A month of empty weeks means the community is a cost center with good vibes.

Where each number actually comes from

None of this needs new tooling in week one, which is worth saying plainly, because "we cannot measure that yet" is the most common reason reporting never starts.

• Finished entry comes from the entry gate itself. If a role is granted at the end of onboarding, the count of role grants is your completion number.
• First answer time comes from support queue timestamps when a queue exists. When it does not, sample ten member questions from the week by hand and record the gap to the first human reply.
• Unanswered past a day comes from one sweep of open threads and support posts, done at the same hour every week.
• Members who spoke comes from the platform's member activity view or a bot export, counted as distinct people rather than messages.
• Moderation load comes from the moderation log, which most teams already keep and nobody reads.
• Decision caused comes from whoever owns the roadmap, in one sentence.

Three columns matter more than the six lines

Here is the difference between a report and a dashboard. Behind every line sit three things, all written before anyone measures anything: an owner, a target, and an if it misses sentence.

Strip those out and you have six facts, which produce a comment in a meeting and no change in the community. Keep them and the page becomes a trigger. The miss is unarguable, because the target was agreed while nobody knew the result. The response is already decided, so the fix does not depend on anyone having spare attention that week. And the owner is a person, not a team, because a number owned by a team is owned by nobody.

Make the response specific enough to schedule. Rewriting two entry steps and retesting with five new joiners is a response. Improving onboarding is a mood.

Targets are commitments, not observations

Choose the target first, then look. Do it the other way round and you will grade yourself against whatever you happen to see, which is how a slow first reply becomes the house standard over two quarters without anyone deciding it should be.

There is no universal right number here, and you should distrust anyone who hands you one. What you sell decides it. A community attached to a paid product owes its members a different response time than a free room for hobbyists. Useful test: pick a first answer time you would be comfortable reading aloud to a paying customer, and an entry completion rate you would defend in front of your board. Date both, and leave them untouched for a quarter.

The meeting that follows

The rule that keeps this alive: anything that hit its target gets zero discussion time. Only misses are discussed, and only the written action is confirmed. A report that fits in five minutes gets sent for years. A thorough one gets sent twice.

Where reporting sits in the wider operation

After a month, the page stops being about numbers and starts naming absences. If you cannot report entry completion, there are no defined entry steps to complete. If you cannot report a first answer time, there is no queue, only channels that certain people happen to watch. If moderation load swings without explanation, escalation paths were never written down.

That is the honest scope. Reporting is one layer, and the rest of the operation includes onboarding and the first two days, response coverage, channel and role architecture, support routing, moderation load and escalation paths, automation coverage, documentation, and the loop that carries member input back into the product. Build the report yourself, this month, without help. Then read what it exposes, because the list of layers you have not built is far more valuable information than any single week's numbers.

What breaks this report

Three failure modes, and all three arrive quietly.

The first is drift toward a highlight reel. Somebody adds a wins section, the wins section grows, and inside two months the misses have been pushed to the bottom of the page where nobody reaches them. Keep the misses at the top, above everything else.

The second is a line owned by a team. Community is not an owner. When a number misses and the owner is a group, the response turns into a conversation about whose job it was, which is how the same miss shows up four weeks in a row with a different explanation each time.

The third is growth. Every curious question asked in a meeting wants to become a new line. Hold at six and answer the curious question once, in the room, rather than by permanently widening the format. A page a busy executive finishes in two minutes gets read for years. One that takes ten minutes gets skipped, and a skipped report is worse than no report, because everybody now believes the community is being watched.

Week one, concretely

Pick three lines: finished entry, first answer time, unanswered past a day. Give each one a person, a target set before you look, and one sentence describing the response. Put it on a single page. Send it on Friday.

Next Friday, send it again with last week beside it. That side by side is the whole point, because one week is noise and two weeks is a direction. Add the remaining three lines in week three, once the habit survives a busy week. Do not redesign it before it has been sent four times.

A community you can only assess by reading it will always be a community you carry personally. Build the instrument, hand each line to a person, decide in advance what a bad week costs, and the room starts improving on weeks you never think about it. More at danieljeong.org