I'm researching this space (and exploring building in it). If you've freelanced: what did your worst late-payment look like — amount, how late? What did you actually send or do, and did it work? And the question I'm most curious about: how long did you put off sending that first reminder, and why? I'll compile the patterns (the wording and timing that actually got people paid) and post them back here next week.
I don't have a useful personal war story here, but one pattern I would separate in the research is whether the first reminder is being asked to do too many jobs.
A reminder after the due date often carries at least three signals at once:
That is why people delay it even when the amount matters. They are not only asking for payment; they are renegotiating the relationship.
The prevention angle is interesting because a pre-due note can be much less loaded. Something like: "Just a quick heads-up that invoice X is due Friday. Do you have everything you need on your side, or is there anything I should resend before then?"
That catches admin blockers without implying distrust. Then if it becomes late, the next message is not the first awkward contact; it is a status follow-up.
So I would probably track three buckets separately:
Those probably need different products. Recovery tools help the first and third. Prevention tools might be stronger for the first two.
I'm curious what convinced you the biggest opportunity is helping freelancers recover overdue payments rather than helping them avoid getting into that situation in the first place.
Did your research point more strongly toward recovery than prevention?
The delay in sending that first reminder often reveals something deeper - you're calculating whether the client relationship is worth the awkwardness vs. the actual revenue impact. Early reminders signal you value the small payment relative to the relationship; long delays signal you've written it off mentally. The pattern might not just be about payment recovery but about how freelancers assess sunk vs. future cost of managing that client.
This is a sharp frame — and it matches about half of what I'm collecting. The relationship-calculus version definitely shows up: one freelancer described waiting 6+ months on a $300 invoice from a long-term client, fully aware she's trading the money for the relationship. But the other half doesn't look like calculation at all — it looks like avoidance: someone described rewriting the same reminder four times over two weeks without ever sending it. No cost-benefit running. Just dread on a loop.
So maybe the delay has two species: a deliberate write-off (your model) and a paralysis spiral (no model at all). One design consequence I find interesting: a heads-up sent before the due date carries zero relationship signal — the calculus never gets to start.
Did you freelance yourself? Curious whether you've ever caught that calculation running live — and what tipped it either way.