Sam Yagan of Corazon Capital on AI's transformative impact on consumer tech and the importance of product-market fit

Summary

Sam Yagan, founder of SparkNotes and Match Group, and now managing a $100M consumer tech fund at Corazon Capital, argues that AI is collapsing the cost of building software fast enough to fundamentally reshape what early-stage investing looks like — but not in the way most founders expect.

Engineering costs vs. marketing costs

The capital efficiency gains from AI hit the engineering side clearly. Yagan says founders will spend far less of their early capital writing code and standing up infrastructure — he reaches back to SparkNotes' early days, when the team had to physically lease and rack servers in colo facilities. Getting an app into the app store is now dramatically cheaper. But he draws a hard line on the marketing side: AI companies haven't figured out how to acquire consumers more cheaply, and word-of-mouth curiosity around new AI products is a thin substitute for a real distribution strategy. The cold-start problem doesn't solve itself just because the product is AI-native.

Product-market fit as the only moat

With engineering and infrastructure costs falling, Yagan argues the real separator among founders becomes a single question: can you build something consumers actually love? That's what he says Corazon knows how to underwrite, and it shapes the fund's thesis directly. The check sizes and stage focus reflect a bet that the scarce resource is product judgment, not technical horsepower.

The COGS problem nobody is pricing in

The sharper structural point is about unit economics. Consumer internet was historically a zero-marginal-cost business — once you built the product, distribution was essentially free. AI breaks that model. Token costs and inference fees create a real cost of goods sold inside consumer tech businesses for the first time. Yagan flags image and video generation products specifically: demand at current pricing may not survive the price levels required to build a real business. He's not worried about this at seed stage — his approach is to find product love first and solve the business model later — but he treats it as a genuine structural shift, not a temporary pricing anomaly.

The SparkNotes pivot and the cheating question

Yagan's own founding history illustrates the pivot thesis he backs. SparkNotes started as a humor site modeled on The Onion before the team recognized it was a bad business and pivoted to study guides. On the ethics of what became a product students used to shortcut homework, Yagan is untroubled: the product gave users the best path to learn the material, not a paper to turn in. The positioning strategy that followed was to get teachers to endorse SparkNotes over CliffsNotes — which some did, explicitly — repositioning a perceived cheating tool as a study aid by making the better product.

AI and education

Corazon is invested in Brilliant, which Yagan describes as a native AI tutor. His framing is that the personalized AI tutor becomes the next generation's equivalent of the graphing calculator — a standard tool for getting through school. He's skeptical of the lazy read that AI has already solved education; the application layer still matters, and the hardware form factor is an open question. The suggestion that a dedicated AI learning device — purpose-built for education rather than trying to replace the phone — could be the next TI-84 is speculative, but it fits the broader thesis that consumer AI products need focused context to win.