1
1 Comment

My tool made people wait a week for value. So nobody came back. Here's how I fixed it.

I'm building SubKitt — it turns your GitHub commits into tweet drafts, so developers who ship constantly but hate posting can stay visible without the marketing chore.

45 days in, I had 0 users. I'd half-convinced myself the product was just too weak to sell.

Then I actually looked at the flow: you connect GitHub, and then... you wait until Monday for your first drafts. A whole week of nothing. For a tool with zero brand yet, that's a death sentence — people connect, feel nothing, and forget it exists by Monday.

The product wasn't too weak. The first-run experience was. The value was a week away.

So I rebuilt it. Now: connect GitHub, and 5 tweet drafts from your actual commits appear in about 60 seconds, each with a one-click post button. It even shows the source commits, so you can see the drafts come from real work, not generic AI noise.

The lesson I keep relearning: when nobody converts, it's tempting to blame the product. Usually it's the path to the value that's broken, not the value itself.

Still 0 paying users — but now the aha-moment is instant instead of a week away.

If you ship code but hate posting about it, it's free during beta: subkitt.com — genuinely want honest feedback, especially on whether the drafts sound human or like noise.

on June 1, 2026
  1. 1

    It's great that you were able to discover that lesson on your own just by thinking through the user journey, without having any actual user feedback yet.

    If I were a Twitter/X user, I'd definitely try testing your service. Personally, it's probably not something I'd use myself, so I'm not really your target audience.

    That said, I'll take a closer look at the landing page later. Maybe I'll spot something useful or come up with a few suggestions that could help.

Trending on Indie Hackers
6 weeks solo, 2 rejections, finally live but nobody told me marketing would be this hard User Avatar 91 comments Building ExpenseSpy solo, no funding — launching June 17 on iOS & Android User Avatar 39 comments Hi IH — quick update. The MVP is live. User Avatar 34 comments I built a $5/1k-listing CRE data API because CoStar is overkill for first-pass scans User Avatar 18 comments Day 7: 51 people answered my question. I wasn't ready for what they said. User Avatar 18 comments Building LinkCover – Day 3: Payment is live. No more building, time to sell. User Avatar 12 comments