01
Why is watch time the more useful number?
Because it measures the clip rather than its distribution. Every viewer had the same opportunity to stop watching, so the average is a property of what you made. Views depend on how many people were shown it, which depends on how the first small group responded.
02
Which one you produce
Views follow from watch time and early response rather than causing them, which is why working on the second moves the first and not the reverse.
| Aspect | Watch time | Views |
|---|---|---|
| Produced by | The clip itself | How widely it was shown |
| Same opportunity for every viewer | Yes | No |
| Comparable between your own clips | Yes | Poorly, distribution varies wildly |
| Responds to improving the clip | Directly | Indirectly, and eventually |
| Useful for deciding what to change | Yes | No |
03
Improving watch time
Start with the opening: get to the point immediately. Remove repetition and slow passages in the middle. End as soon as the point lands. Each of these increases the share of the clip people watch, which is the part of watch time you control.
Clear words on screen also help viewers stay, especially with sound off.
04
Reading the retention graph
The graph shows where viewers leave. Sharp drops point to specific moments that lost people. Watch those moments in your clip and ask what went wrong: a pause, a tangent, a confusing sentence.
Apply the lesson to future clips from similar material.
05
Length and watch time
Longer clips naturally have more room to lose viewers, so their average percentage watched is often lower. Compare clips of similar length to judge fairly.
A long clip with high retention is a strong result.
06
Rewatches
When viewers watch a clip more than once, watch time can exceed the clip length. Clips that loop smoothly or reward a second look encourage this.
Rewatches are a strong sign viewers enjoyed the clip.
07
Why views vary more
Views depend on how widely the platform distributes the clip, which is influenced by many factors outside your control. Watch time reflects the clip itself. Improving watch time tends to lead to more views over time.
Focus effort on the number you influence.
08
Tracking both
Record both numbers for each clip and compare trends over months.
Improving watch time usually lifts views eventually.
A rising median share watched across a month of clips is one of the clearest signs your editing is improving, even before total views catch up.
09
Work on the number you control
- Treat views as the downstream number. It follows from watch time rather than the reverse.
- Compare watch time as a proportion of length. Twenty seconds of a thirty second clip is a different result from twenty of ninety.
- Fix the drop, not the reach. Reach responds to the fix; it does not respond to being wanted.
Make one clip and watch where people leave. Nothing to pay for the first.
Questions people ask
- Can a clip have high views and poor watch time?
- Regularly, usually where a strong opening promised something the rest did not deliver.
- Is a longer clip penalised on watch time?
- It is harder, since more seconds means more chances to leave. The proportion watched is the fairer comparison.
- Which matters more to distribution?
- Watch time and early response, on every major feed. Views are largely a record of what that produced.
- Does rewatching count?
- On several platforms a loop counts again, which is why an ending that leads back to the start is worth building.
- What is a good figure?
- Yours, from last month. Published benchmarks vary too much by format and audience to be worth chasing.