Facebook News Feed

The Problem

PM diagnosis, step one: observe behavior. By 2006, Facebook had a growing real-identity graph, but discovery still looked like a directory workflow. If you wanted to know what friends were doing, you clicked into profiles and scanned the Wall, photos, or recent changes one by one.

PM diagnosis, step two: recognize scaling friction. That pull-based pattern worked for small graphs, but it scaled poorly. Every additional friend increased the number of places you needed to check, which meant the product asked users to do more work just to stay current.

PM diagnosis, step three: notice inefficiency in user effort. The activity already existed across the network, but its value was trapped in scattered places. Users were effectively doing manual aggregation, which is fragile, time-consuming, and easy to abandon.

PM problem statement: How might we turn existing social activity into a single, habit-forming surface that keeps people up to date without forcing them to hunt across profiles?

What Happened

The decision logic was a classic PM bet on shifting behavior: move the homepage from pull to push by aggregating friend activity into a default, continuously updating feed. The hypothesis was that discovery friction, not activity supply, was the primary constraint on habit and return frequency. If the graph could be converted into a living surface, users would come back more often and contribute more because the product felt alive every time they opened it.