
The Job Search Stack
Automate your job search with Claude. Scrape, track, and fol
I'm a product designer. Earlier this year I was job searching and getting nowhere with the manual grind, so I built a system to automate the parts that actually moved the needle: scraping roles into a Google Sheet, tracking them, and drafting a personalized recruiter follow-up after every application. It runs on Claude Cowork, so there's no server and nothing recurring to pay for beyond a subscription I already had.
I put it on Gumroad on a whim after a few friends asked for it. It's at 33 sales at $49 so far. Not life-changing money, but it's real, and it came from something I built in the cracks of my own search.
Two things surprised me, and they might be useful if you're selling something similar.
First, distribution. Every single sale came from LinkedIn. Not one from Reddit, and it wasn't for lack of trying. I posted in the job-search and cs-career subs and got auto-removed or banned from four of them. The honest lesson is match the channel to where your buyer already trusts you. My buyers know me as a designer on LinkedIn. On Reddit I was just another guy with something to sell, and that mod culture treats that as guilty until proven otherwise. I stopped fighting it.
Second, the best marketing turned out to be using the thing myself. Over the last month the system put me in interview rounds with 16 companies and I ended up with 3 offers. I killed about four processes myself once the first offer came in. That result is now the entire pitch. I don't have to argue the tool works, I can just say what it did for me.
The part I didn't expect: a customer in France emailed asking for a non-US version, and building it was nearly free since I was already opening the scrapers to global sources. One email turned into a second product line. Worth listening when a customer tells you what they'd pay for.
Happy to answer anything about the build, the Cowork automation, or the Gumroad side.
About
Built it during my own job search because the stuff that worked didn't scale by hand. So I automated the whole pipeline. Last month it got me 3 offers.

2 Comments
The customer request turning into a second product line is probably the strongest signal here. It’s a good reminder that buyers can reveal adjacent demand you wouldn’t think to build for upfront.