Three hours back on market research
The author spent about a month tracking Reddit by hand to test a product idea. The research showed the people who needed it most could not pay, so the idea stopped before more spend. He then built a weekly system in n8n. The system pulls threads that match set keywords, and an AI sorts each thread as a frustration, question, tip, or idea. Each item gets a short takeaway. The system saves the results and emails a weekly digest.
Start with one job only. Fill in the three lines. Keep the bot on that job.
The one job this bot does: ______________________________
What it may do on its own: ______________________________
What it must ask me first: ______________________________
Two lines from this case, side by side. No new figures.
The author spent about a month tracking Reddit by hand to test a product idea. The research showed the people who needed it most could not pay, so the idea stopped before more spend. He then built a weekly system in n8n. The system pulls threads that match set keywords, and an AI sorts each thread as a frustration, question, tip, or idea. Each item gets a short takeaway. The system saves the results and emails a weekly digest.
The digest saves 3+ hours a week. The AI cost was $12 in a month across six automations, and the other tools run on free tiers or a low-cost n8n plan. One workflow must hold one purpose. Mixing scopes made the results messy, so he clones the workflow for each new scope.
The after line is the source's claim, not our measurement.
smalloffice.space · Get a case like this in your inbox each morning. Subscribe free
Read this case on your phone now, and keep it for later. Type your email. We send this case to that address, and you get the next cases each morning too.