Picking your next book is weirdly hard work. You scroll reviews, ask a friend, check TikTok, add the title to a "want to read" shelf that's already 400 books deep, then end up rereading something you've read twice.
And reading time is getting scarcer. The share of Americans reading for pleasure on an average day has fallen by 40% in 20 years, to about one in six. When you only get a few books a year, a dud costs you more than it used to.
The tools meant to fix this haven't kept up. Goodreads has over 150 million members and is still the default, but since Amazon bought it in 2013 the product has barely moved: whole-star ratings, no dark mode and an iOS app sitting around 3 stars. Sam Parr summed it up when he said he wishes he owned Goodreads. The reviews and community are great. Search, lists and AI are wide open.
Here's the kicker. Most keen readers already have years of taste data sitting in their Goodreads history: every five-star, every abandoned book, every grumpy three-star review. That's an incredibly rich signal, and almost nobody uses it properly. Readers either start from scratch on a new app or get served the same bestseller as everyone else.
Margins, an AI book discovery app that turns your reading history into a personal librarian.
The wedge is deliberately narrow: nail "what should I read next?" for heavy readers who already track. Tracking, reviews and community come along for the ride.
Market size