The case at a glance
- First payment: $70, one customer
- Time building the product before that payment: 8 months
- Total time trying to make a SaaS work: almost 2 years
- Revenue before this payment: $0 (previous product had traffic and signups, zero payments)
- Buyer's stated blockers before he'd pay: 2 (Linear integration, security improvements)
- Time from "yes, I'd pay" to shipped and collected: a few weeks
The product is Announcify, a hosted changelog SaaS companies embed on their own domain so they don't have to build one. GitHub and Linear integrations, AI-generated entries, multiple widget layouts, custom domains, view analytics. Framework and hosting undisclosed.
The operator is solo. His own diagnosis: he was hiding behind development, because coding is easier than asking. The Reddit post telling the story pulled 441 upvotes and 329 comments, and not one of them named the actual asset.
What he's actually selling (it isn't the changelog widget)
The product isn't the changelog tool. The product is two weeks of his own hands, scoped and priced against one specific person.
Here's the sequence. A buyer in Canada installed the widget, used it, and listed what was missing. Instead of building the list and hoping, the founder asked point-blank whether shipping those two things would produce a payment. He got a yes in the same thread, then opened the editor. That's a pre-sale wearing a feature request's clothes.
The mechanism works for a reason most readers will skip: the 8 months of over-building are why the conversation converted. His own lesson — "I should have talked to people sooner" — is only half right. Talk to people early, yes. But the buyer's objections came out as two items instead of twenty because the thing in his hands was already load-bearing. A conversation converts in proportion to how much of the product is already real.
Worth noting what the buyer bought into: an embed, not a dashboard. Once the script sits in someone else's app shell, it stops competing for attention every week. Removing it is a deploy, so nobody bothers.
If you have to guess what the missing piece is, you don't have a product. You have a hobby with a deploy pipeline.