What YouTube recommendations look at

There is no trick to beat the algorithm, but you can understand what it values.

YouTube recommends videos by predicting whether a viewer will be satisfied. Focusing on viewer satisfaction beats any trick.

Signals it uses

• Click-through rate (CTR): how often an impression turns into a click
• Watch time: how long people keep watching after clicking
• Satisfaction: likes and dislikes, survey answers, whether people keep watching
• Viewing history: your video is shown to people who watched similar ones

What you can do

• Show what the video offers within the first 30 seconds
• Cut slow parts without mercy
• Low CTR? Fix the thumbnail and title. Low watch time? Fix the opening.

Where to check

YouTube Studio → Analytics → Content shows impressions click-through rate and average view duration for each video. Compare your best and worst videos side by side.

Check where views come from

YouTube Studio → Analytics → Content → "How viewers find you" shows whether views came from Browse (home), Suggested videos, Search or the Shorts feed.

• Lots of Search: videos that answer what people look up work well for you
• Lots of Suggested: you're showing next to similar bigger videos, so keep going with that topic
• Lots of Browse: thumbnail and title click-through matter most

Next guide: Copyright strikes and the yellow icon: how to avoid them