AN-029Problem Solving

What's a good retention rate for short-form?

Benchmarks age badly — read your finish rate by format against your own baseline, and let loops push the honest number past 100%.

The Diagnosis

Any hard number someone quotes you ('good retention is X%') is stale by the time you hear it and was measured in a different niche anyway. The durable way to read short-form retention: finish rate is the king metric, because on a feed, completion is the strongest satisfied-viewer signal there is — a thousand exits offered, none taken. Judge it by format against your own history, since a 15-second video and a 60-second video are different difficulty settings and comparing them flatters the short one automatically. Then remember loops: a seamless loop legitimately pushes completion past 100% on some platforms because replays count, which is why the seam is worth engineering. And read the curve, not just the rate — WHERE the leaks happen matters more than the average, because a timestamp is fixable and an average is just a mood.

Go Deeper

Related Questions

Solved this one?

The next answer can reach you first.

The handbook grows.

New entries, new answers, free asset drops. No schedule, no spam — just when it ships.