Campus Scout
A housing emergency in the family, turned into a working product and a real business plan
The problem
My brother-in-law had about a month to find off-campus housing near Cal State Fullerton. Listings are scattered across the university marketplace, Craigslist, roommate sites, and Facebook groups. Nothing ranks them against a student’s real constraints, and scammers target students on purpose.
I didn’t start with a pitch deck. I built the thing he needed.
What I built
A nightly scout, built with Claude Code. Every evening at 8pm it swept every source, dropped duplicates, checked that each listing was still live, scored scam risk, ranked by bike distance and budget, and emailed a report with a running shortlist and outreach drafts he could send as-is.
It ran 45 nights straight, through the day classes started. Along the way it caught wrong phone numbers and fake-cheap listings before a real person acted on them. One thing I didn’t expect: the binding constraint wasn’t price. In the final week, about half the new rooms near campus were female-only.
Then I asked if it was a product
A working prototype isn’t a business, so I scoped it like one: competitors, infrastructure costs, and California law. A couple things changed the whole shape of it.
California’s Prepaid Rental Listing Service law makes charging renters for listings a licensed activity, with a DRE license and a $10,000 bond behind it. So listings became the free tier, and the paid tier sells the tools around them (personalization, outreach drafting, scam screening), which the law exempts. The constraint and the growth strategy turned out to be the same design.
The other lever is cost. Researching once per campus and sharing that across every subscriber at the school brings the marginal AI cost to about $0.45 per user against a $19 subscription. The whole stack runs about a dollar a month at zero users.
Decisions I’d defend in an interview
| Decision | The tradeoff |
|---|---|
| Listings free, tools paid | Gave up the obvious revenue line to stay outside a licensing regime, and got a free growth engine in exchange |
| Research per campus, personalize per user | More pipeline complexity for a 5 to 8x drop in marginal cost |
| A $39 season pass that doesn’t renew | Students churn because they signed a lease. Pricing that expires matches how they actually buy, and it sidesteps California auto-renewal law |
| Defer the LLC | California’s $800 a year franchise tax isn’t worth it for a free beta. The triggers for forming one are written down |
| Never scrape | AI search with citations, linking out to the source. Thinner data, but it avoids the pattern that cost Padmapper a $3.1M settlement |
Where it stands
The scout found him a place he could have taken. For reasons outside the search he ended up going a different direction, but he’s housed, and the tool did what it was built to do. It stopped when the housing season did. The scoping report, legal analysis, decision log, and a phased roadmap with exit criteria are all done. If I pick it back up for next year’s cycle, the next step is a landing page and a small free beta at CSUF. Payments, the LLC, and other campuses are deferred on purpose.