Video and YouTube marketing glossary

Audience retention

Audience retention shows how much of a video people actually watch. In YouTube Analytics it appears as a graph plotting the share of viewers still watching at each moment, so you can see exactly where people drop off, skip ahead or rewatch. It is one of the clearest signals of whether a video is holding attention.

Views tell you how many people pressed play. Audience retention tells you what happened next. You will find it in YouTube Studio under the Engagement tab for each individual video. It is a video-level report, so there is no single retention graph for a whole channel.

How to read the graph

The shape of the line is what matters. YouTube describes four patterns:

  • A flat line means people are watching that section from start to finish.
  • A gradual decline is normal. Almost every video loses viewers steadily as it goes on.
  • A spike means more people are watching, rewatching or sharing that moment.
  • A dip means people skipped that section or left altogether.

The first 30 seconds

YouTube reports an Intro figure: the percentage of viewers still watching after 30 seconds. A sharp drop here usually means the opening did not match what the title and thumbnail promised, or it took too long to get going. That is why the hook and the packaging matter as much as the main content.

Retention, average view duration and watch time

These three get mixed up a lot. Audience retention is the graph, moment by moment. Average view duration is one number: the average time watched per view. Watch time is the total hours everyone spent watching. Retention tells you where a video loses people. Average view duration and watch time tell you how much that cost you.

Using it to make better videos

The useful question is not whether retention is high, but where it falls and why. Find the first big dip, watch that moment back and ask what changed. Common causes are a long intro, a section that repeats itself, or a sponsor or product segment placed too early. On a brand channel, comparing the retention graphs of the last ten uploads side by side is often the quickest way to spot a pattern.

YouTube also shows typical retention, which compares a video with your 10 latest videos of a similar length. That is a fairer benchmark than an industry average, because retention varies so much by format and length.

Retention data usually takes one to two days to process, so give a new upload a couple of days before drawing conclusions.

Frequently asked questions

What is a good audience retention rate on YouTube?

There is no single number that applies to every channel. A 60-second Short and a 40-minute interview will never look the same. The most useful comparison is YouTube's own typical retention line, which benchmarks a video against your recent videos of a similar length.

Why does one part of my retention graph spike?

More people watched that moment than the moments around it. Viewers may have rewound to see it again, shared it, or jumped straight to it from a chapter. Spikes are worth studying because they show what your audience finds most valuable, or occasionally what confused them first time.

When does YouTube highlight key moments on the graph?

YouTube only highlights key moments, such as top moments, spikes and dips, on videos that are at least 60 seconds long and have at least 100 views. Retention data usually takes one to two days to process after upload.

How we can help

YouTube Video Production

Related terms

Further reading

Back to the full glossary