Choosing Absolute vs Relative Retention

Two numbers for two questions

Absolute retention answers how many viewers who started are still there at a given time. Relative retention answers how many viewers who reached that time stayed past it. These are different questions, so mixing them leads to wrong edits. Audience retention analytics are clearest when you pick the right one for the section.

For example, a large mid-roll drop looks severe in absolute terms if many left at the start, but relative terms may show it only affected a small committed group. Without that split, you might rewrite the intro instead of trimming a pause.

When to use absolute retention

Use absolute retention for anything tied to the click and the hook. The first 0 to 15 seconds is almost always judged in absolute terms because you care how many of the original viewers entered the video. End-screen performance can also be checked in absolute terms to see total tail loss from all starters.

Also use absolute retention when comparing hooks between videos with similar traffic. It shows the real reach cost of a slow start. If your goal is to raise total watch time, absolute retention at key points is the direct measure to watch.

When to use relative retention

Use relative retention for mid-roll pacing. Once viewers are past the hook, you want to know if that section held them given they were there. This isolates cuts, topic shifts, or pauses from earlier losses. It is the cleanest way to judge edits between minute two and near the end.

Relative retention is also useful for comparing mid-roll sections between videos with very different hook strengths. It levels the playing field so a strong-start video does not make a sloppy middle look fine, and a weak-start video does not make a tight middle look bad.

Applying both in practice

Start at the top: fix absolute retention at the hook first. Then move to the worst interval by relative retention in the middle. Finally check absolute retention at the tail for end-screen losses. This order matches the impact on total watch time for most uploads.

When testing a mid-roll trim, measure relative retention for that exact interval. When testing an intro cut, measure absolute retention for 0 to 15 seconds. Using the wrong metric can make a helpful cut look neutral or a neutral cut look good.

Absolute for starts. Relative for middles.

Avoiding common mix-ups

Do not compare relative retention of a late section to absolute retention of an early section as if they are the same scale. They answer different things. Also do not average relative and absolute values. Keep them separate in your notes and in your tests.

Watch for cases where both move. A good mid-roll trim can lift relative retention in that interval and may also lift absolute retention later, as more committed viewers reach the end. Track both to be sure the change is not a local artifact.

Key takeaways

• Absolute retention: how many of all starters remain at a time point.
• Relative retention: how many of those who reached a time point stayed past it.
• Use absolute for hooks and end screens, relative for mid-roll pacing tests.
• Never mix the two scales when comparing intervals or declaring winners.
• Track both when a fix could affect commitment past the changed interval.