How to Review YouTube Shorts Analytics

A Short can start playing without earning sustained attention. Before deciding that a hook worked, separate starts, continued viewing and the action you wanted a viewer to take. This guide provides a repeatable review for an individual Short and a small group of comparable uploads. It does not offer a target percentage or a formula for recommendations.

Open the correct content view

In YouTube Studio, open Analytics, then Content and the Shorts view. YouTube describes this channel-level view as including views, likes, subscribers, discovery sources and a how-viewers-engaged report. For an individual upload, select that Short and record the report actually available. Do not copy a channel total into a video row. Save the selected dates, export time and whether the figure describes one Short or all Shorts.

Separate a start from continued viewing

YouTube’s current documentation says views count when playback starts. Its engaged-views definition concerns viewers who stayed beyond the initial seconds and excludes loops; stayed-to-watch describes the percentage who continued past the initial seconds of a Short. These measures answer different questions. Record them as separate columns rather than inventing a universal duration threshold. This source check was made on 11 November 2026; a historical export may use an earlier definition.

Use a small worksheet before comparing hooks

Build one row per Short: internal content code, publication date, duration, opening promise, observation window, views, engaged views, stayed-to-watch if shown, and one intended outcome. Keep the raw report labels. A blank metric means unavailable or not collected, not zero. Keep paid promotion in a separate field. If you have bought engagement, identify that activity separately; those counts do not establish organic interest or an audience of prospective customers.

  • Choose one review window, such as each upload’s first seven complete days.
  • Group similar formats and lengths; flag an unusually long or short upload.
  • Write the practical question: clearer opening, better demonstration, or stronger next step.
  • Store a link to your own private report without publishing account or viewer information.

Worked example: two useful but different outcomes

Imagine two fictional product-care Shorts. A has 8,000 starts and 3,200 engaged views. B has 4,000 starts and 2,400 engaged views. A has more of both counts in absolute terms, while B has a larger illustrative engaged-views-to-starts ratio: 60% versus 40%. That arithmetic is an editorial ratio, not YouTube’s stayed-to-watch calculation. Do not rename it as an official metric. If A and B differ in length, distribution or observation window, the ratio is only a prompt for investigation, not evidence that B’s hook caused better results.

Turn a finding into one controlled edit

Watch the opening with the sound off and then with sound. Note where the promised result first becomes clear, whether text can be read before it disappears and whether the demonstration needs context. Choose one change for a future upload: show the finished result before the explanation, for instance. Keep the topic and length similar where feasible. Record what you changed before looking at the next report. If reach differs sharply, state that the comparison cannot isolate the edit’s effect.

Review the next step separately

A view or subscription count does not establish a sale. If the goal is a website visit, check the destination report and any properly configured campaign link separately. A person may watch without clicking, click without a measurable website session, or return later. Do not present an unsupported funnel by multiplying unrelated platform and website totals. At FLR, compare service descriptions and platform risks separately from this editorial content review; paid activity is not a substitute for audience research.

Frequently asked questions

Keep the report’s scope beside every conclusion.

What is a good stayed-to-watch percentage?

This guide provides no universal benchmark. Compare your own similar Shorts over the same observation window and inspect the underlying counts. A percentage alone cannot explain distribution.

Can I calculate stayed-to-watch from engaged views divided by views?

Do not assume so. Use YouTube’s displayed metric and definition; a ratio you calculate from two exported counts should retain its own explicit label.

Should I delete a Short with fewer views?

Low views alone are not a deletion rule. Check whether the information remains useful and accurate, whether the upload had enough observation time and what a revision would test.

Official references and next reading

Definitions above were checked against current YouTube Help on 11 November 2026. The worksheet and fictional comparison are FLR editorial methods, not YouTube instructions or ranking guidance.

YouTube Help: Understand your content performance