In 2009, Kevin Systrom started building an app called Burbn. It was a location check-in app, the kind that was fashionable at the time, and it did a lot. You could check in at places, make plans with friends, earn points for going out and post photos. In March 2010 he closed a $500,000 seed round from Baseline Ventures and Andreessen Horowitz, and Mike Krieger joined him to build it out.

Burbn had a problem: people found it confusing, and it was up against bigger, better-funded check-in apps. Most founders in that spot do one of two things. They add more features to stand out, or they spend the money on marketing and hope. Systrom and Krieger did something harder. They looked at what the people using Burbn were actually doing with it.

Almost nobody was using the check-ins or the plans or the points. What people did use, over and over, was the photo sharing. So they made the uncomfortable call to throw out nearly everything they’d built and keep only that one piece. They rebuilt the app around taking a photo, making it look good with a filter, and sharing it, and nothing else. They named it Instagram.

It launched on October 6, 2010. More than 25,000 people signed up on the first day. It passed one million users on December 12, about two months later. In April 2012, Facebook bought it for roughly $1 billion. The product that sold for a billion dollars was a small fraction of the product the founders had first raised money to build.

What makes this a real turnaround, and not just a lucky launch, is how much it cost to get there. They had to admit that most of a year’s work didn’t matter to the people using it, and delete it. That’s a far harder decision than adding something new, which is why so few people make it.