Grow Creator Field Notes

Viewed vs swiped away on Shorts: what it means

The share of feed viewers who chose to watch your Short instead of scrolling past. Where to find it in Studio, how to read it, how to fix a low one.

Viewed vs. swiped away is the share of people who, once your Short appeared in their feed, chose to watch instead of scrolling past. It grades your cover frame and your first second, nothing else. Retention is a different number on the same page.

YouTube's own description of the metric runs to one sentence. It "highlights the percentage of times that viewers viewed your Shorts versus swiped away", per the Content tab analytics help page for Shorts. That is close to everything the platform has published about it. No target, no benchmark, no guidance for the week it falls off a cliff. The rest of this page is the part Studio leaves out.

Where to find it in Studio

Two routes, and they answer different questions.

Channel level, which tells you whether your packaging is a habit:

1. Open studio.youtube.com and click Analytics in the left rail.
2. Go to the Content tab.
3. Select the Shorts chip in the filter row above the cards.
4. Read the two Shorts feed cards together. Studio labels them "Shown in feed", which YouTube defines as the number of times your Shorts showed in the Shorts feed, and "Viewed (vs swiped away)".

Single Short, which tells you whether one upload was an outlier:

1. Click Content in the left rail.
2. Hover the Short and click the analytics icon.
3. Scroll to the Shorts feed section for that upload on its own.

Read the channel number first. One Short below your usual line is variance. Ten in a row below it is a pattern worth treating as a packaging problem, and packaging problems are the cheap ones to fix.

The confusing part: this is a pre-watch number

Every other metric in Shorts analytics describes people who already decided to stay. Average view duration, average percentage viewed, likes, comments: all of it is measured on the survivors.

Viewed vs. swiped away is measured on the people who never became viewers at all. That difference explains a situation creators write in about often, where a Short holds nearly everyone who starts it and still goes nowhere. The retention graph looks excellent because it only ever described a small group. The reverse happens too. A striking opening frame pulls people in, the middle has no payoff, and a healthy percentage sits next to a view count that stalls early and never recovers.

The metric answers one narrow question. Did the thumb stop? It says nothing about whether the video was worth watching.

Read your five numbers in this order

Checking these out of order is how creators end up rewriting scripts when the problem was distribution, or buying a light when the problem was the first sentence.

1. Shown in feed. If this is small, the percentage beside it is noise computed on a handful of impressions. A weak-looking figure on a low impression count is not a verdict on your packaging. The likely explanations are that the Short is new, or that its traffic arrived through your subscriptions feed, your channel page, search or an external link rather than through the Shorts feed itself.

2. Viewed vs. swiped away. With real impressions behind it, the number means something. This grades the opening frame plus the first instant of motion and sound.

3. Views against engaged views. Since 31 March 2025, a view counts every time a Short starts to play or replay, with no minimum watch time, and the older definition survives in Analytics under the name engaged views (YouTube Help). The gap between the two is a separate diagnosis: people who stopped scrolling, started the video, then left inside the opening beat. A wide gap points at your first spoken line rather than your cover frame.

4. Average view duration and average percentage viewed. These describe the middle. If checks 2 and 3 look fine and this one collapses, the problem is structure, usually a promise made in the hook that the body never pays off. Reading those two together is covered in our Shorts retention benchmark guide.

5. Average view duration set against the length of the Short. If the average comes out longer than the video, replays are the explanation that fits, since a single pass through cannot exceed the runtime. Treat it as the closest thing to a rewatch signal Studio hands you without extra arithmetic.

A reading guide, not a benchmark

YouTube publishes no target for this metric. Creators report figures on both sides of half depending on niche, Short length and how much of their traffic comes from subscribers, so an absolute pass mark taken from another blog is somebody's guess wearing a lab coat. The ruler with no guesswork in it is your own history.

Build it once. Open the Content tab, select the Shorts chip, set the date range wide enough to cover your last 20 Shorts, note the viewed vs. swiped away figure for each, and take the middle value. That is your median.

Direction matters more than level. A channel whose median climbs over six weeks is fixing something real, whatever the raw percentage happens to be.

Why the number reads differently since the 2025 view change

Two things changed on 31 March 2025. Views on Shorts began counting every play or replay with no minimum watch time, and the previous definition stayed available as engaged views. YouTube states on the same help page that Partner Program eligibility and Shorts ad revenue sharing are both based on engaged views, so earnings did not move with the relabelling.

The knock-on effect is a comparison problem. A Short from 2024 and a Short from this year do not carry comparable view counts, which means any personal benchmark you built before that date is measuring a different thing. YouTube has not published how the redefinition interacts with the viewed vs. swiped away card, so rebuild your median from recent uploads only and stop comparing across that date.

The confusion the change created is worth naming. Creators see a respectable view count sitting next to a poor viewed vs. swiped away percentage and assume one of the two is broken. Both can be honest at once, because the view count counts plays while the swipe metric counts decisions. A Short can collect plenty of starts from a large pile of impressions and still be passed over by most of the feed.

Fixing it, in the order the feed sees it

The opening frame. In the Shorts feed the video starts playing on its own, so what a scrolling viewer sees is your first frame rather than a thumbnail. A custom Shorts thumbnail still matters on your channel page and in search results and is worth setting, but it is not what the feed puts in front of the swipe. Look at your first frame at thumb size with the sound off. If it shows you sitting down, adjusting the camera, or an empty background before anything happens, that frame is doing no work. Cut to the moment where the interesting thing is already on screen. For a second opinion on the frame itself, the free reel cover analyzer reads a cover image the way a scrolling viewer does, in roughly the glance it gets.

The first spoken line. Greetings and setup sentences spend the exact moment you are being judged. Say the specific thing first. "Hey everyone, so today I wanted to talk about lighting" uses two seconds to convey nothing, while "your ring light is why your footage looks cheap" uses the same two seconds to make a claim someone can disagree with.

The first on-screen text. Read it aloud. If saying it takes longer than the opening beat of the video, most people deciding whether to stay will not finish it. Large enough to read at arm's length and specific enough to be worth reading.

A made-up example, one idea packaged twice:

• Before. Opening frame: a wide shot of an empty desk with someone walking into it. On-screen text: "Day 47 of my journey". First line: "Hey everyone, welcome back to the channel."
• After. Opening frame: the finished lighting shot already lit and on screen. On-screen text: "Hardware store lighting setup". First line: "This replaced the expensive softbox I had been carrying around for a year."

Same footage, same information, different first second. More hook patterns that survive that second are collected in short-form hooks that stop the scroll.

Instagram does not give you this number

Instagram Insights has no equivalent card as the app stands today. A Reel gives you views, accounts reached, watch time, average watch time and interactions, and none of those separate the people who scrolled past from the people who never had it served to them. No impressions-to-plays ratio is exposed in the app, so there is no Instagram version of viewed vs. swiped away to go looking for.

The practical workaround is cross-posting. Publish the same cut to both, let YouTube tell you whether the packaging works, then apply the answer to the Reel. The differences between the two analytics panels are laid out in Reels vs Shorts analytics.

What to do this week

1. Build your median from your last 20 Shorts and write it somewhere you will see it again.
2. Pull the three Shorts furthest below that median and put their opening frames side by side. The shared habit is usually obvious within seconds.
3. Post the next three Shorts changing only the opening frame. Same topics, same edit, same posting time.
4. Compare those three against the median, not against each other and not against your best Short ever.
5. If the median moves, keep the change. If it does not, the failure is at check 3 or check 4, and the next thing to rewrite is the first spoken line.

The other cards on that page, and what each one points at, are walked through in how to read your Shorts analytics.

Canonical: https://growcreator.pro/blog/viewed-vs-swiped-away-youtube-shorts

Leumos Private Limited · hello@growcreator.pro