YouTube Counts a View From the First Frame.

From 24 August YouTube counts a view the moment a video starts. Public numbers will jump, engaged views will not, and your reporting needs both.

On 17 August YouTube announced that from 24 August a public view is counted the moment a video starts playing.

First frame, no minimum watch time, on long form video and live streams as well as Shorts.

The old watch-time-based number survives under a new name, Engaged views, inside YouTube Analytics.

So public view counts are about to rise without a single extra person watching anything. If your monthly report carries one line called “views”, it is about to show growth you did not earn, cannot repeat and cannot explain to the person reading it.

What changes on 24 August

YouTube has never published the exact threshold a play had to clear before it counted. The working assumption across the industry has been around 30 seconds of watch time on a long form video.

From 24 August that threshold is gone. A view registers at the first frame of playback, and it registers when someone enters a live stream.

Shorts already worked this way, having moved to play-based counting in 2025. The change puts every YouTube format on one definition, and brings YouTube into line with how TikTok and Instagram already count.

YouTube’s stated reason is that running several counting systems across formats created confusion, and that creators want a number reflecting their true exposure.

Engaged views is the number that did not move

The watch-time metric has not been deleted. It has been renamed and relocated.

Engaged views sits in YouTube Analytics under Advanced Mode, and it counts the people who stayed past the opening seconds rather than everyone whose player started.

That matters more than the headline. Engaged views is the continuity number, the one you can still compare to July, and to last year, without an asterisk.

YouTube has also confirmed that earnings and Partner Programme eligibility are unaffected. Payouts continue to run on engaged Shorts views and engaged watch hours, not on the public counter.

Why the next report will look like growth

Two numbers now describe the same video: how many players started, and how many people stayed.

Public views will be the larger of the two, and on some videos it will be a lot larger. A channel whose traffic leans on browse and suggested placements, where autoplay does much of the starting, will see a wider gap than a channel fed by subscribers and search.

None of that is a distribution change. The algorithm did not move, the audience did not grow, and nothing about the content got better on 24 August. Only the counting rule changed.

A metric definition change is an operations event

This is the part most teams get wrong. They file a redefinition with the feature announcements, as platform news, when it actually lands on reporting, targets, contracts and client expectations at the same time.

Platform numbers get redefined regularly. Instagram removed inactive accounts in May 2026 and follower counts fell overnight, which is why the metrics worth reporting are the ones that survive a definition change rather than the ones that flatter you this month.

A team that has written down what each number in its report means absorbs this in an afternoon. A team that has not spends the next quarter arguing about whether it grew.

What to do before the report goes out

Five moves, and none of them are heavy:

  • Split the view line in two. Public views and engaged views, labelled, side by side. A single number invites the wrong question.
  • Export August engaged views for every active video now, so you hold a clean pre-change baseline you control rather than one you have to reconstruct later.
  • Add a one-line footnote to the report for the next three months: public views count from the first frame from 24 August, engaged views do not.
  • Re-read any target, bonus or client deliverable written as a view count and decide which definition it means. Put that in writing before someone else decides for you.
  • Recalculate anything derived from views. Retention rate and view-to-subscriber ratios especially, because the denominator just got bigger.

Then check the rest of the report for the same weakness. Working through a structured audit of what you measure and why usually turns up two or three other numbers nobody in the room can define either.

The comparability break, and how long it lasts

August 2026 will not compare cleanly to July 2026 on public views. Nor will the twelve months after it compare to the twelve before.

YouTube has said the new metric applies from 24 August onwards wherever views are shown, and it has not said it will restate past totals. So the archive stays as it is and the break sits in the middle of your series rather than being smoothed away.

The practical answer: run year on year comparisons on engaged views, and treat public views as a reach line rather than a performance line until you have twelve clean months of it.

What this does not change

Worth naming, because a rising number tempts people to relax.

  • Watch time. The minutes people actually spend are untouched, and that is still what compounds on YouTube.
  • Earnings and Partner Programme eligibility. Both continue to run on engaged measures.
  • Whether the video was any good. A first-frame view can be someone leaving.

The risk here is quiet. A team that starts optimising for the public counter is optimising for plays started, which rewards packaging and does nothing for retention. Packaging earns the click. Retention earns the next recommendation.

What good looks like

A report where every number has a written definition, a source, an owner and a date it last changed.

That reads like bureaucracy right up until a platform moves. A brand with that document gets a note the week of the change, updates one line, and sends the report on time with a footnote. A brand without one finds out in the leadership meeting, in front of the person who signs off budget.

The second version costs credibility, and credibility is what a social team spends when it asks for more resource.

How NBK thinks about metric changes

NBK treats measurement as part of the operating system, not as a tool someone opens at month end. Definitions, owners, cadence and the escalation path for a platform change belong in the same place as the approvals and the publishing calendar.

Which is also why the reporting rhythm matters more than the dashboard. Numbers move for reasons that have nothing to do with your work, and a team reporting monthly with context absorbs those moves without drama. Reporting results to leadership is far easier when every line was defined long before the meeting.

Organic growth is judged over quarters. Any counting change is noise until you can show what it did to the trend.

Next step

Before your August report goes out, open it and try to define every number on it in one sentence. Anywhere you cannot, that number is a liability rather than a measure.

If your social output feels busy but not effective, an NBK audit is a sensible place to start.

Written by Matt Cunnelly, edited to the NBK Social editorial standards. AI-assisted research and drafting, human-edited and fact-checked. Spot an error? Tell us.

Matt Cunnelly, Founder & CEO, NBK Social. 15+ years building social for global publishers, from UNILAD (LADbible Group) to Supercar Blondie (SB Media). Focused on the systems behind consistent, large-scale growth.

Newsletter

The NBK Social briefing

Our YouTube coverage, and everything else we publish, by email.

Free · Unsubscribe in one click
Subscribe