YouTube views jump 76% while ad revenue crashes 55%

By Dominic Ashworth · 2026-07-30

Views are up. The cheque is down.

Australian creators and brand managers feeling good about their YouTube view counts should look at a different number. The figure that pays the bills has been heading the wrong way all year. Metricool's analysis found average long-form views per video rose 76% to 5,985 between February 2025 and February 2026, yet estimated ad revenue fell 55% over exactly the same period.verifiedVerified Source: metricool.com Under any normal logic of how advertising works, those two lines should not move in opposite directions.

The figures come from a Metricool analysis of 799,718 YouTube videos drawn from 71,177 accounts worldwide, published in July 2026.verifiedVerified Source: metricool.com Blanca Ruiz, the study's author, said the scope was broad by design: the analysis covered videos from accounts across every industry, niche, and account size to find out what is really happening on YouTube in 2026. Bushletter could not independently verify the underlying data, which comes from Metricool's own platform and proprietary research.

Watch time is the missing link

The mechanism is straightforward, even if it is easy to miss when you only look at view counts. YouTube places mid-roll advertisements inside videos that run beyond a minimum duration threshold. If a viewer clicks away before that threshold, no mid-roll runs, no impression is recorded, no monetised playback is logged, and no revenue is generated regardless of how many people started watching.

Average view duration across the Metricool sample fell 37% to 2.51 minutes in February 2026, compared with February 2025.verifiedVerified Source: metricool.com At two and a half minutes, a large share of viewers are gone before any mid-roll slot opens. A channel can look busier than ever while delivering fewer and fewer billable moments to the advertiser on the other side of the transaction.

Ad impressions declined 51% year-on-year, and monetised playbacks dropped 44% across the same period. A creator who doubled their audience over the past year could still be collecting a smaller cheque today than twelve months ago, and the view count column gives no warning of that at all.

How the platform got here

YouTube's own architecture contributed to the dynamic. Ruiz said no other platform puts hours-long content and videos just a few seconds long in front of the same audience, on the same channel, in the same app. That breadth is a competitive strength, but it also trains viewer behaviour in ways that do not always serve long-form creators.

Shorts and algorithmic recommendations pull users toward shorter clips. When those habits bleed into how people consume long-form content, clicking in, sampling a minute or two, then moving on, the mid-roll economy breaks down. Viewers who arrive on a 20-minute video through a recommendation feed bring habits formed on platforms where stopping after 90 seconds is entirely normal. The content may be long-form; the viewing session often is not.

What brands buying creator media should ask

Australian marketing teams allocating budget to creator partnerships or running pre-roll and mid-roll campaigns face a direct implication from the Metricool data. A creator quoting subscriber counts and total views in a media kit is presenting numbers that, on current evidence, can sit entirely disconnected from actual ad delivery. The questions worth asking are about average view duration, audience retention curves, and monetised playback rates, not headline view counts.

The same logic applies to in-house brand channels. A company measuring its YouTube investment by views per video is running a vanity metric as its primary KPI. If viewers leave before the mid-point of a ten-minute brand video, the content investment is generating impressions only at the very start, where no mid-roll can run, and nothing after that. The watch-time report inside YouTube Studio is the document that reveals whether a content strategy is actually working.

There is also a structural risk for creators who have built sponsorship rates around historical view averages. If average duration has compressed by 37% and that compression is not visible to the brand paying for an integration in the back half of a video, the effective reach of that placement has shrunk without any corresponding adjustment to the price. Brands that start requesting duration data before signing deals will find the market reprices itself quickly.

Which metrics to watch instead

View duration is the first number to fix in any YouTube reporting dashboard. Alongside it, audience retention percentage at the 30%, 50% and 70% marks shows whether viewers are reaching mid-roll windows at all. Monetised playbacks, available in YouTube's revenue reports, show directly how many sessions generated an ad serving event. Impressions per view is a derived figure worth calculating: divide total ad impressions by total views and track whether it holds steady or keeps falling.

For brands evaluating creator partnerships, asking for a screenshot of the YouTube Studio retention graph for recent comparable videos costs nothing and reveals more than any media kit statistic. A creator whose audience watches 60% of a fifteen-minute video is a different commercial proposition from one whose audience watches 15%, regardless of what either channel's subscriber count says.

The Metricool study is vendor-produced research and should be read accordingly. The mechanics it describes, mid-roll ad slots that require minimum watch time to serve and a platform ecosystem that is shortening average sessions, are structural features of how YouTube's advertising model works. The numbers may shift as viewing habits evolve, but the underlying logic does not. If watch time keeps falling, revenue will keep following it down, no matter how strong the view count looks on the surface.

Dominic Ashworth

Dominic Ashworth writes about marketing, media and brands. He is fascinated by how attention is won and lost, and by the mastheads and campaigns that manage to hold it.