View rate past the first 3 seconds: what viewed vs swiped away means

What the phrase usually means

People who ask about the view rate past the first 3 seconds want to know one thing: of everyone who was shown the clip, how many stayed past the opening. The platforms do not all use that wording. This guide sticks to what YouTube has said in its own pages, because those are the pages we could read and check.

The phrase is also searched with Instagram attached. We did not find an Instagram Help Center page that defines a metric with this name in this run, so we do not describe Instagram's version here or quote a threshold for it. If you see the figure in your own Instagram insights, read the label and the info text next to it in the app.

What YouTube says about Viewed vs swiped away

The only text we found is a video post on YouTube's Community help forum dated 8 May 2024. It is community content on a YouTube Help page, not a Help Center article, and the page marks community content as possibly not verified or up to date. It says YouTube is introducing two new Shorts metrics in YouTube Analytics, Shown in feed and Viewed (vs. swiped away).

• What it shows: how often a Short was shown in the Shorts feed, and whether viewers chose to view it or swipe away, alongside the existing view count.
• Why it was added: the post says the earlier thumbnail impression rate could not tell you whether a Short was shown in a feed or viewed or swiped away.
• Where to find it: for all Shorts creators, on Studio web and Studio mobile, in Advanced Analytics and in Content-level Analytics.

Related numbers that are defined

YouTube says to use Engaged views to compare performance across your Shorts. That is the cleanest official guidance on which number to line your clips up by. Because a plain view now counts at the start of playback, a view count alone no longer tells you that anyone stayed.

Retention has its own report. YouTube explains the graph this way: a slow decline is normal, spikes mean viewers are watching, rewatching or sharing, and dips mean they are skipping or leaving. Its key-moments version needs a video of at least 60 seconds and at least 100 views, and the page does not say how that applies to shorter Shorts.

How to read a batch of clips

1. Put clips from the same episode side by side. They share a guest, a topic and an audience, so differences are more likely to come from the opening.
2. Line them up by the same measure. Use Stayed to watch or Engaged views for the comparison, and keep Viewed vs swiped away as a second view of the same start.
3. Look for the outlier before the average. One clip with a clearly weaker start is more useful to you than a channel-wide number.
4. Write down what the first line was. A table of clip, first sentence and result teaches you more than the result alone.
5. Wait until each clip has been shown a fair amount. YouTube gives no minimum sample, so we cannot offer a number; treat early readings as a first look.

For ideas on what to put in that first line, see TikTok hook ideas and hooks for YouTube Shorts.

How to test two openings of the same moment

The cleanest test changes one thing. Pick a moment that works, then cut it twice: the same end point and the same captions, but a different first sentence. Post the two versions on different days and write down which opened each. Because one moment is the same story both times, any gap points at the opening rather than the topic.

• Keep length, captions and audio the same in both versions.
• Post them on different days and record the date and time of each, so you can see if timing is muddying the comparison.
• Run the test on more than one moment before you decide a pattern is better. One result can be luck.

KeyMo is useful at the front of this process. It cuts up to 15 vertical clips from a video, scores each out of 10 with a one-line reason, and keeps only those scoring 7 or higher. To make the alternate opening, open the clip in the chat editor and tell KeyMo what to change. It will not tell you which opening wins; your analytics do that. See pricing for plans.