The Idea

Kyle Fowler had been collecting cards since he was a kid, and like every collector he had a shoebox problem. He wanted two simple things: to know what he actually owned, and to know what it was worth. So in the winter of 2019 he went looking for an app that could digitise his baseball collection and identify a card from a photo.

It didn't exist. Not in the form he wanted, anyway.

Kyle was a self-taught developer. He wasn't scanning the market for a hot niche or writing a thesis about the trading card economy. He just knew, with the specificity only an obsessive has, exactly what was annoying about his own hobby. Every collector he knew had the same two questions and nobody had built the answer.

So he built the thing he wanted. Point your phone at a card, get the name, the set, the rarity and the market value, and watch your collection add itself up. Then he posted about it on Reddit and waited. The downloads started coming. This was the beginning of Cardstock.

The Execution

• Winter 2019: Kyle started building Cardstock alone, for a reason he later wrote into the App Store listing himself: he couldn't find a good app to catalogue his baseball collection. No team, no funding, no market research.
• 2020: It took roughly six months to reach a working version, including a three-month gap in the middle when it stalled entirely. His first downloads came after he posted an update on Reddit, which is the whole distribution strategy: find the room where the obsessives already are.
• 2020 to 2025: Cardstock compounded quietly into a real business, reaching around 15,000 active subscriptions, roughly $75,000 in monthly recurring revenue and about 13,000 downloads a month. Still a solo operation.
• 2025: Instead of chasing a new market, Kyle ran the identical playbook on the biggest card hobby on earth. He built the Scanémon MVP in a single day using Cursor's agent capabilities. The first app took six months. The second took a day.
• 2025 to 2026: Within roughly a year of launch, Scanémon hit around 3,400 active subscriptions and about $17,000 MRR, pulling in approximately 16,000 new customers a month. Pokémon collectors had the same shoebox problem baseball collectors did.
• 2026: Scanémon 2.0 shipped graded card scanning for PSA, BGS and CGC slabs, condition-based pricing and a rebuilt set tracker, every feature driven by collector requests. There is no free tier. The entire app sits behind a subscription and collectors pay it anyway.
• Today: The two apps clear seven figures a year between them, and Starter Story featured Kyle as the self-taught developer who turned a hobby into one successful app, then did it again.

The lesson?

Kyle didn't find his idea by scanning the market for a hot niche. He found it by asking what was annoying about a hobby he already couldn't shut up about, which is a question only he could answer with that much precision. Domain obsession isn't a soft advantage, it's the moat. Note the timeline too: the first app took six months of nights and weekends, the second took a day, because by then he knew exactly which problem to point the tools at. AI collapsed the build time. It didn't hand him the insight. Everyone is hunting for the perfect idea. Kyle just fixed the thing that was already irritating him.