All Reports

Everyone wants to spy on ads. I'd rather get paid when a card declines.

Nov 6, 2026

Daily dashboard screenshot

Everyone wants to spy on ads. I'd rather get paid when a card declines.

Three things jumped out of the 74 signals today, and only one of them made me open a notes file.

The top scorer is Track competitors' winning ad creatives at 74. The pitch is good because the pain is real: you're dumping $4k/mo into Meta and you genuinely don't know if the guy two clicks down your feed is running the same angle with a better hook. Monetization is $49–$199/mo tiered by ad volume.

Here's my problem with it: the free Meta Ad Library already exists, and on top of it you've got Foreplay, Atria, AdSpy, BigSpy, and about nine others I keep seeing in the same Twitter threads. The scrape isn't the moat, the tagging is, and everyone's tagging is mid. Churn will be brutal because agencies sign up for a month when a client launches and cancel after. If I built it, I'd skip the dashboard entirely. Just a weekly email: "5 new creatives in your niche that have been running 14+ days" at $29/mo. Ads that survive two weeks are the actual winners — everything else is noise. Small product, low churn, way easier to sell. But I probably wouldn't build it. Too crowded, and I don't want to fight a scraping arms race with Meta.

The one I'd actually build is Failed-payment recovery and smart dunning for Stripe subscriptions, 73 points, $29–$79/mo flat or 10–15% of recovered revenue.

Yes, Stripe has Smart Retries baked in. No, it doesn't write the email sequence, it doesn't pause access at the right moment, it doesn't ping you in Slack when a $400/mo account goes into dunning, and it definitely doesn't do the win-back campaign three weeks later. Involuntary churn is the quietest revenue leak a SaaS has — you don't get a churn survey, you get a card that expired. I've been on the wrong end of this and it stings.

The rev-share pricing is a trap, by the way. Nobody wants to hand over 15% of recovered revenue forever, and it makes your revenue unpredictable. Flat $49/mo with a "we recovered $X this month" number on the dashboard — that's the whole sales pitch. Distribution matters more than the product here, and the wedge is obvious: ship it as a Stripe App. The Stripe App Marketplace is one of the few remaining channels where a solo dev can get in front of thousands of businesses with actual payment intent already expressed. Competition is Churnkey and the free tiers of Baremetrics and ProfitWell, which is real. But those bundle it into a bigger analytics product. A sharp, single-purpose tool at $49 might still win the small end of the market.

Third, and this one is weak: Plain-English statistics for Excel and CSV files, 69 points, free tier plus $12–$29/mo.

The demand is real — "just tell me the average and whether the weird row matters" is a genuine need for ops people and analysts stuck in spreadsheets. But willingness to pay is near zero because ChatGPT does this for free if you paste the data. The only wedge is local processing, no upload, saved repeatable reports. Privacy angle. That's a $12/mo product at best with a nasty CAC problem. I'd skip it.

The Ship AI agents to production without touching infra idea at 73 is a great pitch and a terrible market for a solo dev. Vercel, Modal, Cloudflare, Fly, Railway, and every YC batch is on this. You can't out-infra them.

Stack if I were building the dunning one tomorrow: Next.js on Vercel, Supabase for state, Stripe webhooks + a couple of cron jobs on Trigger.dev, Resend for the emails, and Resend again for the onboarding. That's a weekend to something chargeable. Maybe ten days to a real Stripe App listing.

The pattern I keep noticing: the highest-scoring ideas are the ones with the loudest pitch, not the widest gap. Ad spy tools shout. Dunning tools whisper, and the whisper is attached to money that's already moving.

Back to all daily reports