The most boring idea on today's list is the only one with real money attached

16 new cards today, 42 signals, and 14 of the 16 cards came from the same source (competitor_gap). Zero signals from any platform feed. Which tells you something: today isn't about what's trending, it's about what's missing. Fine by me. Missing features are easier to charge for than trending ones.

The thing I actually stopped on: Automated failed-payment recovery for Paddle and Lemon Squeezy merchants. 76 score, highest of the day. The pitch is dumb-sounding — "we email people when their card expires" — but the gap is real. Every dunning tool on the planet points at Stripe: Baremetrics, Churn Bounty, ProfitWell Retain, the whole graveyard. Paddle and Lemon Squeezy merchants get whatever dunning their MoR ships with and nothing else. And these are exactly the people who picked an MoR so they'd never have to think about billing ops.

Would it make money? Yes, and that's rare for me to say on a Tuesday. Failed payments are a pure subtraction problem — you can point at the recovered invoice and say "that's yours, I took 10%." Pricing at $29–99/mo tiered by MRR is sane, but honestly the success fee is the wedge. I'd launch with only the success fee for the first 20 customers, because "$0 until you get paid" closes a solo founder in one email. Then convert them to flat monthly once they trust it.

The catch: it's a support-ticket business, not a ship-and-forget business. Every recovery flow breaks differently per merchant, per currency, per bank. You're signing up to debug other people's billing. I'd still build it — Next.js, Paddle webhooks into Supabase, a Vercel cron for the retry ladder, Resend for the emails. Two weeks of work. But I'd cap it at the two MoRs and refuse every "can you also do Stripe?" email, because that's where the 20 competitors are waiting.

Then there's Unified inbox for scattered customer messages at 73. I get why it scores — the pain is visceral, everyone hates tab-switching between email, DMs and the site form. But this is a meat grinder. Front, Missive, Chatwoot, Intercom, and about forty dead startups are already in it, and the $19–49/mo per-seat model means your revenue is capped by the customer's headcount, which for a small team is like three. I wouldn't build it. The only version that survives is vertical: inbox for Shopify sellers, inbox for course creators, inbox for agencies. Generic is a death sentence here.

Upload a spreadsheet, ask a question, get answers at 65 is the one I'm most skeptical of, and it's the one that'll get the most clicks. Eleven months ago this was a company. Now it's a weekend. $12–19/mo, free tier of 5–10 analyses — that pricing tells you everything, it's consumer-grade churn with a business-grade support burden. The LLM part is twenty lines. The hard part is explaining to someone why the answer to "how many customers churned in Q3" is wrong because their sheet had merged cells. I'd skip it. If you're going to do spreadsheets, do the boring ops version: reconcile these two exports and tell me what doesn't match.

Stack of choice for anything this week: Next.js on Vercel, Supabase, Resend, and a cron job. I keep waiting for that to stop being the answer and it keeps being the answer.

The thing rattling around my head: 14 of 16 cards came from looking at what competitors aren't doing. That's a fine hunting ground, but it's also how you end up building version 1.4 of somebody else's roadmap. The failed-payment one is the only card today where the gap is big enough to stand in.