We were looking at video views, which is the worst metric for managing a video website.

In hindsight, we didn't track revenue effectively; we focused on usage metrics, which made it unclear how many monthly active users we needed for fundraising.

In the early days, we tracked visitors, leads, and customers, but we should have focused on churn rates to understand customer satisfaction better.

Chasing subscriber counts can be misleading; they don't always correlate with views or engagement, which are more important metrics.

A bug in our caching system caused the model to respond as if it were talking to itself, leading to a significant drop in user satisfaction until we identified and fixed it.

We can't manage what we can't measure, so we're starting with measurement, but it's not sufficient on its own.

We had a beautiful growth graph, but we realized we were running a business with only a 5% margin, which isn't sustainable for a software company.

When product teams focus solely on metrics like engagement or dwell time, they risk neglecting qualitative aspects such as content quality and user satisfaction, which can lead to long-term retention issues.

As Twitch scaled, we focused 70-80% of our engineering efforts on improving our video system because if the video didn't work, it didn't matter how good the rest of our product was.