YouTube Shorts: views vs engaged views, and which one belongs in your report

YouTube changed what a Shorts "view" means on 31 March 2025. If your reporting still treats the number as continuous across that date, your growth chart contains a step that has nothing to do with performance.

The two metrics

Views. Every time a Short starts — first play or replay — it counts. No minimum watch time. This is the number on the public counter and the one most dashboards read.

Engaged views. Plays where the viewer stayed past the opening seconds or interacted. This is effectively the old view definition, and it is the metric tied to monetisation and Partner Programme eligibility. It sits in Advanced mode in YouTube Analytics rather than on the summary screen.

Both are legitimate. They answer different questions:

The step change nobody annotates

Because the new definition includes every replay and every one-frame swipe-past, reported view counts jumped on 31 March 2025 without any change in audience behaviour. A year-over-year deck built on total views will show a growth spike straddling that date. Two ways to handle it:

1. 1Use engaged views for anything crossing the boundary. They map onto the old definition, so the series stays honest.
2. 2Start the series after the change and annotate the chart. Fewer data points, no false narrative.

What you should not do is average the two, or quietly switch metric mid-series.

What this does to cross-platform comparisons

After the change, all three major short-form platforms count a view at playback start. That sounds like it makes them comparable. It does not, quite:

• Autoplay context differs. A Shorts view can be generated by a swipe that lands on the video for a fraction of a second, and the same is true on TikTok. How aggressively each feed autoplays is not something you can normalise away.
• Replays are included everywhere, but loop behaviour differs by format and length. Short loops inflate more.
• Only YouTube gives you the stricter metric. There is no "engaged views" equivalent published for Reels or TikTok, so a like-for-like quality comparison across platforms is not available from public counters alone.

The practical answer is to compare platforms on trend and on ratios, not on raw view totals — engagement per view, share of views earned in the first 48 hours, cost per thousand within one platform. See cross-platform reporting for the full method.

If you pay creators on Shorts views

Name the metric in the agreement. "Views as shown on the public video page" and "engaged views as reported in YouTube Analytics Advanced mode" can differ substantially on the same video, and the second requires channel access the brand may not have.

For campaigns where the brand cannot see the creator's analytics, public views are the only auditable number — so use them, but set the rate with the knowledge that they include replays and shallow plays. A rate built for the old definition applied to the new one silently overpays.

Checklist before your next Shorts report

• State the metric name and definition in the report header.
• Flag any series that crosses 31 March 2025.
• Use engaged views for monetisation forecasts and historic comparisons.
• Use the delta between two timestamped readings, not the lifetime counter, for campaign attribution.
• Keep the raw readings. When a number is challenged three months later, the snapshot history is the only thing that settles it.