YouTube Changed How Views Are Counted: Do Your Numbers Still Mean Anything?

You opened Studio one morning and the numbers were bigger. Not one video, everything, including uploads you had not touched in months. Your audience did not double overnight, so either something broke or something moved.

Something moved. YouTube changed what counts as a view, and the creator side of the platform has been arguing about it ever since.

The short answer

Since 24 August 2026, YouTube counts a view the moment a video starts to play, on every format: Shorts, long-form and live streams. The stricter count did not disappear. It is called engaged views, it counts the people who stayed past the first seconds, and it is still what your earnings run on. So your public numbers went up while your audience and your revenue stayed exactly where they were. Source: YouTube's own help centre.

What actually changed, in plain terms

Before, a view was recorded once someone had watched enough of your video for YouTube to count it as a real watch. Now it is recorded when playback begins. A video that starts playing while somebody scrolls past it counts. A click followed by an instant exit counts. The bar moved from "stayed" to "started".

Two consequences follow, and only one of them is obvious. The obvious one: your public count inflates. The less obvious one: it does not inflate by the same amount everywhere. A channel whose videos mostly get opened on purpose, from search or from a subscription feed, will barely notice. A channel that lives on feeds, autoplay or Shorts sees a much wider gap between what it used to count and what it counts now. There is no universal multiplier to apply, and two channels of the same size can move by very different amounts. Which is a decent reminder that comparing raw view counts between channels was always shakier than it looked.

One thing did not move at all. YouTube says the update does not affect Partner Program earnings or eligibility: earnings still run on engaged views and engaged watch hours, and eligibility on qualified views. If your revenue graph is flat while your view graph jumped, nothing is broken.

The number that still tells you the truth about your own videos

For your own channel the answer is short: read engaged views in YouTube Analytics. YouTube defines them as the number of times viewers stayed to watch past the initial seconds, loops excluded. That is the old bar, still measured, still there. If your instincts were built on the previous count, engaged views is where those instincts still work.

What you lose is the at-a-glance reflex. The big number on the watch page and the number that means something are no longer the same number, so a quick look tells you less than it did in July. Annoying, not fatal.

The harder problem: reading someone else's channel

This is the part that matters when you are deciding what to make next, and nobody at YouTube is going to solve it for you. You cannot see anyone else's engaged views. On another creator's video you get one number, the public one, and it is now measured with a looser ruler than it was a month ago.

So how do you tell whether a video in your niche actually worked?

The same way you should have been judging it before: relatively. A view count on its own never meant much. Forty thousand views is a triumph on a channel that usually does two thousand and a disappointment on one that usually does four hundred thousand. What carries information is the gap between a video and the ordinary numbers of the channel that published it. A gap is a ratio, and a ratio mostly survives a change of unit. If the video and its channel's baseline are both measured with the same looser ruler, a video doing eight times its channel's usual numbers is still doing eight times its channel's usual numbers.

There is one caveat, and it is real for the next few weeks. The ratio only holds when both sides sit on the same side of 24 August. A video from June compared against a baseline built from September uploads mixes two counting rules, and the recent one will look better than it is. Give a channel a handful of uploads after the change before you trust the comparison, and stay sceptical of anything that straddles the date.

What to change this week

Three things worth doing while the dust settles.

• Raise any absolute view threshold you set before late August. A "minimum 20,000 views" filter written in July is a looser filter today than it was then, and it will wave through videos you would have skipped.
• Judge your own uploads on engaged views rather than the watch page number. Same metric you were already using, different place to look for it.
• For everyone else's videos, read the gap instead of the number. Compare a video to what its own channel has been doing lately, over a window that sits entirely after the change.

Where the tool fits

That last point is the one job NicheFinder does. It runs inside YouTube, scrolls a results page to the end on its own, and for every video computes what you would otherwise check tab by tab: the size and age of the channel, the median views of its recent uploads, and how far this video beats that median. That multiple is the Outlier Score, and because it is a ratio between two numbers counted the same way, it holds up through the change in a way an absolute threshold does not.

The setting that does need your attention is the minimum views filter, which is absolute and yours to control. If you set it months ago, push it up. You can run a search from the homepage without an account: type a niche you work in, look at what is punching above its own channel, and compare that reading with the raw view counts you would have trusted in the spring. If you would rather start from what is already moving, the weekly niche radar tracks which searches are climbing.

The deeper habit has not changed either. A small channel breaks out on the clarity of its promise, not on its production budget, and that is still perfectly readable from the outside.

Frequently asked questions

Did my revenue change?

No. YouTube states that the update does not affect Partner Program earnings or eligibility. Earnings stay based on engaged views and engaged watch hours, and eligibility on qualified views. A bigger public number puts nothing extra in your pocket.

Are the new views fake?

They are real playbacks. They are just not proof that anyone paid attention, which is exactly why engaged views still exists as a separate number. The definition changed, not reality: a play now counts, where before it had to turn into a watch. The public number measures reach, engaged views measure interest.

Can I compare a video from July with one from September?

Not directly. They were counted under two different rules, and the September number carries an advantage that has nothing to do with the video itself. Compare inside a single period, or wait until the channel you are studying has enough recent uploads to give you a clean baseline.

Should I lower my expectations for my own videos?

Adjust them rather than lower them. Engaged views tell you what your audience actually did, and that has not moved. What moved is the number strangers see under your thumbnail, and a slightly bigger number there is not the worst thing for a small channel trying to earn a first click. Picking the right lane still matters more than either metric.

Before your next round of research: check the outliers in your niche with the new counting in mind, and raise the view thresholds you set back in the summer.