How the retention part works
No black box. Here is exactly what goes into the numbers, what they weigh, and what the club is asked to do about them.
What we measure about a member
Out of 100. Attendance carries the most weight because it moves first, but it isn't the whole picture — a member who attends every week and hasn't had a real conversation with anyone in a year is a resignation waiting to surprise you.
| When they last attended, and how often | 40 pts |
| Committees and projects they're part of | 25 pts |
| When someone last had a conversation with them | 20 pts |
| Whether their dues are current | 10 pts |
| How long they've been in the club | 5 pts |
What we measure about a club
Also out of 100. The last line is the one clubs find most useful: it asks whether you acted on what you already knew.
| Attendance, and whether it's rising or falling | 25 pts |
| Members gained and lost, relative to club size | 20 pts |
| Prospective members currently in conversation | 15 pts |
| How many members are on a committee or a project | 20 pts |
| Dues that are current | 10 pts |
| Whether at-risk members have actually been contacted | 10 pts |
What you get on Monday
At most 7 things, ranked. Not a dashboard — a short list where each item names one person, one reason, and one thing to do. A guest who visited and heard nothing back outranks almost everything else, because that window closes fastest.
We deliberately don't send more. A list of forty is a backlog, and a backlog gets closed rather than worked. Everything under the cut is still on the member's own page whenever you want it.
What it will never do
- Email a member on its own. Every message is drafted for a person to read, edit and send.
- Score anyone with a model. The rules are fixed and published on this page.
- Tell a club it is failing. It reports what it measured and what it doesn't know.
- Share a member's name outside the club. District rollups are counts; cross-club sharing strips anything identifying before it leaves.
Questions
- Does AI decide who is at risk?
- No. The score is a fixed set of rules over facts your club already records — attendance, involvement, when someone was last spoken to, whether dues are current. The same facts always produce the same score. AI can help draft a note afterwards, but a person sends it.
- Will it flag our honorary members?
- No. Honorary and corporate members aren't held to weekly attendance, because nobody expects them to attend weekly. Judging them by that standard would fill the list with people who are exactly where everyone agreed they'd be.
- What about someone on a leave of absence?
- Never flagged. They told you they'd be away. Treating that as drifting would generate exactly the outreach they asked you not to send.
- We just joined and everything looks bad. Is our club failing?
- No — we just don't know yet. A club with nothing recorded shows as "not enough information", not "at risk". New members get 90 days to settle before anyone is flagged.
- Can I see why someone was flagged?
- Always. Every score shows its drivers — what was measured, what it scored, and what was available. If you disagree, you dismiss it and say why, and that's recorded too.