Closing the book
Every match ends with an answer. Every answer makes the next match better.
When a match is over, either of you records how it went. That's how you get your next match, and it's how the matching learns what actually leads to good working relationships.
What counts as an outcome
| Item | Outcomes | Follow-up |
|---|---|---|
| Hiring | Hired, or closed | Check-ins at 30 days, 90 days and a year |
| Project work | Awarded and completed, or closed | On time, on budget, quality, would you work together again |
| Mentoring, peers, trade partners | Working together, good contact, not a fit, flaked, ghosted | A check-in at two months |
| Co-founders | "We're partners", or the networking outcomes | Check-ins at 1, 3, 6 and 12 months |
| Circles | Recommit every three months, or leave | A private rating after each meeting |
What happens when you close
A short private review
Why it closed (if the outcome doesn't say), did they match their profile, how they treated the people around you, and whether anything felt unsafe or unfair. Skippable. Never shown to the other person.
Never told which
The other person never learns which outcome you picked. "Not a fit", "flaked" and "ghosted" mean the two of you are never matched again.
Flaking counts only as a pattern
One person's report changes nothing. It takes reports from more than one person before anyone's matches slow down.
Observed beats claimed
What people observe at meetups gradually outweighs what members claim about themselves.
A nudge, not an accusation
If several people flag a detail of your profile as out of date, you get a neutral prompt to fix it. It never says anyone flagged anything.
What's limiting your matches
The app can tell you which of your settings narrows your matches most, like a minimum pay or a schedule. Never a protected or identity answer.
How the matching learns
Each month, for hiring and networking, the app looks at which parts of a match went with good outcomes among people who actually met: a confirmed hire, a second meeting, working together. It then nudges how much each part counts, in small steps and within fixed limits.
Before any update goes live, it's tested: the app re-ranks real matches under the old and new settings and compares groups with the 4/5ths rule. An update that would make any group worse off below that line is blocked and never applied. Every applied update is logged.
No protected trait, age, gender or audit answer is ever read by the learning. More on fairness
Keep exploring
Your first match could be this week.
Free to join. Answer the basics in a few minutes and you can be matched right away. Everything else you add makes matching better.