YouTube engagement is measured against views, not subscribers

Put the video's view count into the followers field and enter likes and comments to get the rate creators actually compare on YouTube.

The calculation happens in your browser, so you can check a private or unlisted video's numbers without connecting anything.

Subscribers are the worst denominator on YouTube. A channel with 100,000 subscribers may show a video 20,000 times or two million times depending on whether browse and search picked it up, and subscribers who never open the notification tab still sit in the count. Interactions divided by views is the number that compares meaningfully between videos.

Read on views, healthy engagement usually lands somewhere between 2 and 6%, with likes making up the bulk of it. Comments run far lower than on the short-form platforms because leaving one means stopping a ten-minute video to type, so a comment on YouTube is worth several times what a comment costs a viewer elsewhere.

The metric that actually drives a video, and that no engagement formula includes, is average view duration. YouTube optimises for watch time, so a video holding people for six minutes out of ten will keep being recommended even with an unremarkable like ratio, while a video with excellent engagement and a sharp drop at thirty seconds quietly stops being served.

Shorts and long-form videos should never be averaged together. Shorts behave like TikTok, pulling enormous view counts with proportionally tiny interaction counts, so mixing them into one channel-level engagement rate will make your long-form videos look far worse than they are. Calculate the two formats separately.