Two screenshots can show different numbers for the same Short without either being wrong. One might display public views, while the other shows engaged views inside Analytics. Before explaining a performance change, identify which number each screenshot represents.
YouTube's August 12, 2026 explanation defines qualified Shorts views for YPP as engaged views from public Shorts. It excludes categories such as private, unlisted, deleted, and ad-viewed Shorts. Engaged views reflect viewers staying beyond the initial seconds and exclude loops. YouTube's qualified-views explanation
That clarification makes this a useful time to rebuild a reporting sheet that has accumulated incompatible metrics.
Label the metric before calculating a rate
YouTube changed public Shorts view counting on March 31, 2025, to include starts and replays without a minimum watch-time requirement. The earlier measurement continues as engaged views. A chart spanning that change needs the definition change marked clearly. TeamYouTube's view-count announcement
Create separate fields for public views, engaged views, reporting period, and the account's eligibility progress where relevant. Do not paste whichever number is easiest to find into a column simply labeled “views.”
For an illustrative internal report, a Short might show 18,000 public views and 9,000 engaged views. Those invented numbers demonstrate the distinction; they are not a benchmark or a prediction. Dividing one by the other produces a ratio you have defined, not automatically a standard YouTube metric.
Keep each comparison answerable
Write the question above the table. “Did viewers continue watching this topic?” requires different evidence from “How many people encountered the upload?” or “What does the Earn page currently show?”
Use a reporting structure like this:
| Field | Why it belongs |
|---|---|
| Video and publication date | Identifies the exact upload |
| Observation window | Makes the comparison reproducible |
| Public views | Records the displayed reach measure |
| Engaged views | Keeps continued watching distinct |
| Relevant retention notes | Adds context about the viewing experience |
| Creative hypothesis | States what the upload was testing |
Record missing data as unavailable. A blank field is more honest than a zero when the report has not provided the metric.
Compare videos at a similar stage
A Short published yesterday and one published six weeks ago have had different opportunities to find viewers. Choose a consistent observation point for your editorial review, while remembering that later performance can still change.
Keep unlike subjects separate when they serve different purposes. A product announcement and a complete tutorial may both be Shorts, but the audience's reason to watch can differ substantially.
Our recommendation is to review small groups with a shared subject, length range, and intended audience. Record exceptions in notes. The result is still observational evidence, not a controlled experiment, but it becomes easier to interpret.
Turn the report into an editing decision
Imagine three hypothetical repair Shorts. Each starts with the same greeting, but only one immediately shows the broken component. If that video's viewing pattern looks stronger, the next test could move the visible problem into the opening.
Do not conclude that removing greetings universally improves performance. State the narrower hypothesis: “For this repair series, viewers may need to see the failure before hearing the setup.”
Write the next upload's opening before closing the analytics review. Otherwise, a careful report can become a weekly ritual that changes nothing about production.
Preserve the history without mixing definitions
Keep the original export and add a note explaining any renamed column or revised formula. When someone opens the sheet later, they should be able to reconstruct what a number meant on the date it was recorded.
For monetization questions, consult the actual Earn interface and the applicable program rules instead of estimating eligibility from a public counter. For creative decisions, use the correctly labeled audience evidence to choose one test.
The deliverable is a report a teammate can read without asking what “views” means. That removes a surprisingly common source of bad conclusions before the next video is filmed.