The AI Video Auto-Editor Making $108k/mo Is Real — And The Other Two Ideas Are Traps

8 new cards today, 156 signals scraped. Most of it is noise. But three things made me stop.

Let's start with the obvious winner: AI-Powered Short Video Auto-Editing Tool. 88 score, $108k MRR, pricing at $19-99/mo. This isn't a hypothetical — someone's already doing $1.3M ARR. The signal is clear: creators are drowning in long-form content and need short clips for TikTok/Reels/Shorts. Every podcaster, every YouTuber, every webinar host has this pain. The monetization makes sense — usage-based with a cap, so heavy users pay more. I'd build this with FFmpeg + Whisper for transcription + some basic scene detection. Ship a simple web app where you upload a video, it spits out 5 short clips with captions. Don't over-engineer. The $19 tier is your hook, $99 is where the money lives.

Now the On-Demand Branded Merchandise Platform at 90 — highest score today. Sounds sexy. "Brands need custom merch fast." But here's the thing: this is a logistics nightmare. You're not just writing code — you're dealing with printing partners, inventory, shipping, returns. That's a business, not a SaaS. Unless you're Printful or Printify, stay the fuck away. The commission model sounds good until you realize margins are thin and support is hell. Hard pass.

The Build AI Chatbot Based on Documents at 88 is interesting but crowded as fuck. Chatbase, SiteGPT, DocsBot — they're everywhere. The differentiation would need to be razor-sharp. Maybe focus on a specific vertical (legal docs, medical manuals) instead of generic "upload PDF get chatbot." But honestly? I'd skip this unless you have a distribution advantage. The $108k video editing tool is a better bet — less competition, clearer pain.

What gets me is the platform distribution: ProductHunt has 986 signals, competitor_discovery 426, payment_discovery 378. Payment_discovery is where the real money signals live — people are literally paying for these things. That's how I found the video editor. The $108k number came from payment data, not speculation.

If I were building today: I'd clone the video auto-editor but target a niche — maybe just podcasters. Charge $29/mo for 10 hours of video processing. Use Replicate for the AI stuff, Supabase for backend, Stripe for payments. Ship in 2 weeks. That's it.

Productivity (824) and developer-tools (494) dominate the category distribution. Everyone's building the same shit. The video tool sits in "AI" (104) — smaller pond, bigger fish.

Real talk: most of these 156 signals are noise. The branded merch thing will trap someone into 6 months of supply chain headaches. The no-code platforms (appearing twice at 88 and 85) are a race to the bottom. But that video editor? That's a real business. Go build it before someone else does.