I found my last startup idea by accident, and I'm not sure it counts
I spent about 8 hours last week trying to find a startup idea the "right" way—talking to people, noticing friction, looking at industries I knew nothing about. I got nowhere useful. Then on Thursday I was fixing a broken Zapier integration at 11 PM and realized I was doing the same manual workaround I'd mentioned to exactly zero people as a problem worth solving. That's the idea I've been poking at since then.
Thing is, I don't know if this counts. The advice usually says ideas should solve something bigger than your own convenience, and I can't tell if "yet another internal workflow tool" qualifies. I've spent maybe $47 on tools this month too — Clipchamp, because someone on Reddit called it essential for quick edits, but I still haven't opened it for this project specifically.
Has anyone tried building something they uncovered by accident rather than through systematic hunting? And did it feel flimsy at first but actually hold up, or am I chasing a workaround that only matters to me?
I actually think accidental ideas can be stronger than “idea hunting” because they come from behavior, not theory.
The real question is not whether the problem came from your own convenience. It is whether your workaround is repeated by a specific group of people who have the same broken workflow.
So I would not validate this as “is my idea big enough?”
I’d validate it as:
Who else is fixing this same Zapier-style break manually?
How often does it happen?
What does it cost them when it breaks?
Are they already duct-taping it with Notion, spreadsheets, scripts, or manual checks?
If the answer is yes, then it is not “yet another internal workflow tool.” It is a workflow reliability problem hiding inside an annoying manual workaround.
The best next step is probably not building more. It is writing the exact broken workflow in one sentence and finding 10 people who already depend on that type of automation.
This counts more than the "systematic" ideas, not less. You already did the thing idea-validation can't buy — you revealed demand with your own behavior. Doing a manual workaround at 11 PM without ever calling it a "problem" is a stronger signal than anyone telling you in an interview that they'd want something. Stated demand is cheap; revealed behavior isn't.
"Bigger than my convenience" is the wrong filter. The real question is n=1 vs n=many: are you the only one quietly doing this workaround, or are there 3–5 others hacking around the same gap and also not calling it a problem out loud? Find even a handful and it's real. Find nobody and — yeah, maybe it's just your itch.
The most useful thing I've built started exactly this way: I forked a tool I was already using and built the pieces I kept wishing it had. Being your own first honest user is a big part of why those hold up. (Also, the $47 of unopened tools is the most relatable line in the post.) What's the workaround itself — Zapier-specific, or a pattern you'd hit anywhere?