Two Metrics, One Underlying Event
Both numbers are built from the same raw material: how far each viewer got before they left. They differ only in what they divide it by. Average view duration reports the result in seconds. Average percentage viewed divides it by the runtime and reports a proportion.
That sounds like a distinction without a difference until you notice that one of them changes when you edit your video and the other changes when you change its length. Cut thirty seconds off the end of a video nobody was watching anyway and the percentage jumps while the duration barely moves. The video did not get better. The denominator got smaller.
This is the trap. Percentage is the more flattering metric to optimize and the easier one to move for reasons that have nothing to do with the quality of the video, which is exactly why it should never be read on its own.
YouTube Treats Them as Separate Signals
This is not a distinction invented by analytics writers. YouTube names both in its own description of how Shorts are ranked, stating that "Our systems use the signals for % of viewers who chose to view, avg. view duration and avg. % viewed to inform ranking."
Three signals, listed separately, describing three different things: how many people agreed to watch, how long they stayed, and how much of the video that was. If percentage were simply duration expressed differently, there would be no reason to name both.
The reason both matter is that they fail in opposite directions. A long video can hold people for a long time and still show a poor percentage, which tells the system the video was longer than its material justified. A very short video can show an excellent percentage while holding nobody for long, which tells the system rather less than the number suggests.
How to Read Them Together
The combination is more informative than either alone, and there are only four cases worth recognizing.
High duration with high percentage is the straightforward win: the video held people, and it was the right length for what it had to say. High duration with low percentage means the material was strong but the runtime outran it, and the fix is in the edit rather than the idea. Low duration with high percentage usually means the video was very short, and the percentage is flattering something that did not hold anyone long enough to matter. Low on both is the honest failure, and it is almost always an opening problem rather than an ending one.
Notice that only one of those four cases is diagnosed by looking at a single number. The others all require both, which is the practical argument for never quoting one without the other. Which number answers which question is worth keeping to hand.
| What you are asking | Read this | Why |
|---|---|---|
| How long did people stay? | Average view duration | Reported in time, not a proportion |
| Did the video earn its length? | Average percentage viewed | A proportion of the runtime |
| Did trimming it actually help? | Both | Cutting lifts one and can lower the other |
| Which does ranking use? | Both | YouTube names them separately for Shorts |
The Length Question Hiding Underneath
Most disagreements between these two metrics are really a question about length, and it is a question worth asking deliberately rather than answering by habit. The right runtime is the shortest one that delivers the payoff without damaging it, and almost nobody finds that by instinct on the first edit.
The honest test is whether every section earns the next few seconds of the viewer's time. A video that passes that test will show a healthy percentage naturally, because there is nothing in it that was not pulling its weight. A video that fails it can be rescued either by cutting the sections that do not earn their place or by giving them something to do.
That is a judgement about content, not about numbers, and it is the point at which analytics stop being able to help. Retensis reads the video itself against the curve, so you can see which stretch of the edit is the one costing you both numbers at once. We also cover how to choose a length in the first place, which is the version of this problem you get to solve before you post rather than after.
Frequently asked questions
One is measured in time and the other is measured as a proportion of your runtime. Average view duration answers how long the typical viewer stayed. Average percentage viewed answers how much of the video that represents. A 90 second video and a 20 second video can have the same duration figure and wildly different percentages, or the reverse, and neither number is more correct than the other.
Yes, and it names them separately. In its guidance on Shorts discovery, YouTube states that its systems use the share of viewers who chose to view, average view duration and average percentage viewed to inform ranking. Listing both rather than one is a strong hint that they are not redundant, and that a video which performs well on one and badly on the other is telling the system something mixed.
Whichever one your content strategy depends on. If you are trying to hold a wide audience for as long as possible, duration is the honest scoreboard. If you are trying to make a tight, complete video that earns its length, percentage is. The mistake is improving one by quietly sacrificing the other, which is what happens when a working video gets trimmed purely to lift the ratio.
On formats that loop it can, because a viewer who watches through and starts again has consumed more than one runtime. That is worth knowing before you celebrate: a figure above 100% is a replay signal rather than an impossible number, and it usually means the video is short and the ending sends people back to the beginning. It also means the percentage has stopped being a clean measure of whether the video earned its length.
Get deeper insight into your YouTube Shorts
Analyze your Shorts before you publish: hook score, predicted retention, and specific fixes. Find out what your YouTube Studio data can't tell you.
Analyze your Shorts free →