Average view percentage is the retention metric YouTube weighs most heavily, and it is largely decided by decisions made before recording: topic, format, length, and structure. The Watch Time Predictor estimates average watch time percentage from those decisions, marks the drop-off risk points by section, gives retention improvement tips, and compares the prediction with your niche average.
This guide covers what drives the prediction, how to read the section risks, and how to use it to choose between two versions of the same video.
What drives the prediction
Length relative to topic (a 20-minute video on a 6-minute topic will not hold), format (tutorial, comparison, vlog, commentary each have characteristic retention shapes), structure (payoff early or late, number of sections, presence of a roadmap), and topic specificity. The predictor combines these with niche norms to produce an estimated average percentage and a confidence range.
Section risks
Each section in your structure is rated for drop-off risk: high for a long setup, a method-before-result section or a late sponsor; low for a result, a reveal or a step the viewer came for. The section list is a redesign guide — reorder so the high-risk sections are shorter and sit after payoffs, not before them.
Choosing between versions
The most useful way to use the predictor is comparatively: describe two structures for the same video (a 12-minute version with method first, a 9-minute version with results first) and compare. A five-point predicted difference is a real reason to pick one over the other before you have spent a day filming.
Tips
- Describe the structure as a numbered list of sections with rough durations.
- Enter your channel's typical average percentage if you know it; the niche comparison becomes a channel comparison.
- Re-run after the Retention Script Analyzer's rewrites to see the effect.
Common mistakes
- Predicting from a topic alone — Structure is most of the prediction.
- Chasing a high percentage with a very short video — A 3-minute video at 70% is 2.1 minutes; a 10-minute video at 50% is 5.
- Ignoring the confidence range — A prediction of 48% ± 8 is a shape, not a promise.
Worked example
Input: "Comparison; 11 min; sections: intro 1:00, methodology 1:30, grinder 1–5 (1:20 each), results 1:30, verdict 0:40. Home espresso."
What the report returned: Predicted 41% (±7); niche average 44%. High risk: methodology (before any result), grinder 4–5 (fatigue). Tips: open with the surprise result, cut methodology to 30s, group grinders 3–5 into one segment. Restructured version predicted 51%.
Frequently asked questions
What is a good average view percentage on YouTube?
It depends on length: 50–60% is strong for 8–12 minutes; 40% is respectable for 20 minutes. The niche comparison gives your context.
How accurate is the prediction?
Reliable for comparing structures and identifying risks; the absolute number has a stated range.
Does it work for Shorts?
It is built for long-form; the Shorts Loop Completion Tracker covers Shorts retention.
What changes retention the most?
Payoff placement — moving the result before the method is the single biggest structural lever.