How-To Guide
How do you read YouTube Shorts analytics to decide what to post next?
Read YouTube Shorts analytics in seven steps: what each Studio number means, which popular targets have no source, and a weekly log to pick your next Short.
By the Superdirector editorial team. YouTube Help pages and vendor guides read November 7, 2026; updated November 12, 2026.
At a glance
How do you read YouTube Shorts analytics to decide what to post next?
To decide what to post next from YouTube Shorts analytics, open YouTube Studio, choose Analytics, then Content, and select the Shorts chip. Then sort each recent Short into repeat, remake or drop by comparing it with your own last ten Shorts of the same type, because YouTube's Help pages give no target for any Shorts metric. Remake it when stay (how many chose to view) or depth (average percentage viewed) falls clearly below your usual, and change only the part that failed: the opening for stay, the middle or payoff for depth. Repeat it when both hold. When only reach (shown in feed) is low, keep the format and post it again on another topic or week, because low reach says little about the video. Drop the idea when the remake also falls short on the same read.
- Compare reach, stay and depth with your last ten Shorts of the same type; stay and depth decide a remake, reach alone never does.
- YouTube's Help pages give no target for any Shorts number; ranges quoted by vendors are their own or creator-reported.
- Views count every start or replay since 31 March 2025; engaged views count stays and exclude loops.
Built from YouTube's definitions of shown in feed, how many chose to view and average percentage viewed. Original diagram; the 0.90 line, the three-day wait and the 500-engaged-view floor are this guide's working rules, not YouTube numbers.
Make it work for your brand
Find it. Remake it. Post it. Superdirector turns the viral formats in your niche into your brand's carousels, X posts and video scripts, and keeps the calendar full.
The short answer: repeat, remake or drop
The short version: sort every recent Short into repeat, remake or drop, and judge it only against your own recent Shorts of the same type. YouTube's Help pages publish no target for any Shorts number, so a stay rate that is weak for a talking-head tip can be normal for a product demo.
Three reads feed that verdict. Reach asks whether YouTube showed the Short to enough people to judge it. Stay asks whether those people chose to watch instead of swiping on. Depth asks whether the viewers who stayed kept going. Studio reports each one separately, and they carry different weight: a weak stay number points at the opening, a weak depth number points at the middle or the payoff, and a weak reach number says little about the video, so reach alone never calls for a remake. The two rules YouTube itself gives for this report are to compare Shorts of one type with each other and to avoid measuring everything against a viral hit; both are on its Content tab tips for Shorts.
The rest of this guide shows where each number lives in Studio, what YouTube says it means, which popular "good number" claims have no source, and a weekly routine that takes about fifteen minutes, and how to turn the verdicts into next week's posts. Our own additions are the repeat, remake and drop labels, the 0.90 line for "clearly below", the minimum sample before a verdict and the one-remake rule, offered as a working method and not as anything YouTube states.
Open the Shorts report in YouTube Studio
- Open YouTube Studio and pick Analytics from the left menu.
- Choose the Content tab, then the Shorts chip. The cards switch to Shorts only: views, likes, subscribers, shown in feed, how many chose to view and how viewers find your Shorts.
- For a single Short, pick Content in the left menu, point at the Short and choose Analytics. Its retention graph sits under the Overview and Engagement tabs, as YouTube's audience retention page describes.
- Use See more or Advanced mode when you want the expanded report to compare Shorts or export the data.
Three notes from YouTube's Help pages prevent false alarms. Studio's Analytics page has been getting a redesign since July 2026, as YouTube's Get started with YouTube Analytics page says, so a card may sit somewhere other than in an older tutorial's screenshot. YouTube says retention data typically takes one to two days to process, and other numbers can take hours to appear, so a Short published this morning has no verdict yet. And YouTube may slow, freeze or revise a count while it discards low-quality playbacks, so a figure you wrote down on Monday can differ on Friday.
Read each number as YouTube defines it
The table paraphrases YouTube's definitions from its content performance page and its Shorts tips page. Read the third column first, because most misreadings start with assuming a number says more than it does.
| Number in Studio | What YouTube says it is | What it cannot tell you |
|---|---|---|
| Shown in feed | How often the Short appeared in the Shorts feed. | How many different people saw it, or whether anyone watched. |
| How many chose to view (also labelled how viewers engaged, or stayed to watch) | The share of appearances in which the viewer stayed past the initial seconds instead of swiping on. This is the stay read. | Whether anyone reached the end, or exactly how many seconds count as a stay: YouTube's pages say only "initial seconds" and give no number. |
| Engaged views | The count of those stays, with loops left out. | Unique viewers, or how many finished. It is the count behind the stay rate, not a depth number. |
| Views | Since 31 March 2025, each start or replay counts, however briefly it plays. | Attention. It only measures exposure. |
| Average view duration | Average time watched, among viewers who stayed, worked out from engaged views and their watch time. | What everyone shown the Short did. |
| Average percentage viewed | Average share of the Short watched, again among viewers who stayed. | How many finished, or how many left early. |
| Subscribers | Change in subscribers, set against your typical performance. | A conversion rate, or whether those subscribers return. |
| How viewers find your Shorts | A split by source: the Shorts feed, search, channel pages, browse features, outside sites and others. | Why one source sent more people than another. |
So the three reads map to these cards. Reach is shown in feed. Stay is how many chose to view, with engaged views as its count. Depth is average percentage viewed and average view duration, with the retention graph for detail. One consequence is easy to miss. The two depth numbers describe only the viewers who stayed, so a Short can lose most people at the start and still show a strong average, because the smaller group that stayed watched a lot. ContentStudio makes the same point and adds a caution: a few repeat viewers can inflate an average that rests on few engaged views, so look at the count behind the percentage.
Match your pattern to one change
Put the three reads side by side as ratios: the Short's number divided by the middle value of your last ten Shorts of the same type, so 1.0 is your usual. Our working line is 0.90. A ratio below 0.90 is clearly below usual; anything from 0.90 up counts as holding. Stay and depth decide whether to remake. Reach only tells you whether the Short got a fair test. The rows are our reading of the numbers, built from YouTube's definitions and the diagnostic tables on practitioner pages, so treat each as a hypothesis to test.
| Pattern against your last ten | Verdict | Likely cause and the one change to test |
|---|---|---|
| Reach, stay and depth all hold | Repeat | Make the next Short on the same premise or format, changing the example. |
| Stay and depth hold, reach is low | Repeat the format on another topic or week | Thin distribution, not a weak video. YouTube lists topic interest, competition and seasonality as outside factors, and says strong metrics can still earn fewer impressions when other channels do better. Compare traffic sources and similar uploads, then post the format again on a different topic or in a different week. |
| Stay is low, depth holds | Remake the opening | The first frame, first line or promise did not earn the stay. Show the result or the problem at once, say plainly who the Short is for, and leave the rest alone. |
| Stay holds, depth is low | Remake the middle | The opening works and the middle or payoff loses people. Move the payoff earlier or cut setup, using the retention graph to find the dip. |
| Stay and depth are both low | Remake the opening first | The topic, audience or execution may be off. Rebuild the opening on a narrower topic; the stay fix comes first because depth only covers those who stayed. |
| Stay and depth hold, subscribers are few | Repeat, with a link to your channel | The Short works alone but gives no reason to follow. Tie it to the subject your channel keeps returning to. |
| A remake is still low on the read it was meant to fix | Drop | Two honest tries, the original and one remake, both failed on the same read. Retire the idea and spend the slot on a new one. |
When reach and stay are both low, follow the stay row: the opening is the part you control. A remake that fixes stay but shows a new weakness in depth is not a failed remake; it earns one more remake aimed at the middle.
Two view-count changes that break old comparisons
If your log spans these dates, mark them. A raw view count from before 31 March 2025 does not measure the same thing as one from after it, and the August 2026 change extended the same idea to other formats.
| Date | What YouTube changed | What stayed the same |
|---|---|---|
| 31 March 2025 | A Shorts view now counts each start or replay with no minimum watch time. The earlier measure, which counted viewers who chose to keep watching, was kept and renamed engaged views. Source: Get started creating YouTube Shorts. | Engaged views remain in Analytics, and YouTube said Partner Program eligibility and Shorts revenue sharing would not change: both rest on engaged Shorts views. |
| 24 August 2026 | A view registers as soon as playback begins in every format: Shorts, long videos and live streams. Source: YouTube's content performance page. | YouTube says Partner Program earnings and eligibility are unaffected: earnings still rest on engaged views and engaged watch hours, and eligibility on qualified views. |
The practical rule is to compare engaged views with engaged views and to write the date next to any change in your log. YouTube's How engagement metrics are counted page explains how views are validated over time.
Numbers you will read, sorted by who stands behind them
Search for "good Shorts numbers" and confident tables appear. We read the pages that rank and checked each claim against YouTube's own pages.
How we checked: on 7 November 2026 we read eleven YouTube Help articles, among them the Shorts Content tab tips, the content performance glossary, audience retention, how Shorts are discovered and how engagement metrics are counted, and searched each for a target, benchmark or threshold. None gives a target for the stay rate, retention or subscribers. The one range on those pages sits in the impressions FAQ, which says half of all channels and videos have an impressions click-through rate between 2 and 10 percent; that is not one of the Shorts cards. Satura and 1of10, both vendors, also say YouTube publishes no Shorts target. In the table, the Status column gives each claim's weight: YouTube-defined, vendor target, creator-reported or in conflict with YouTube.
| Claim you will read | Who says it | Status |
|---|---|---|
| A 70 to 79 percent stay rate is "working" and 80 or more is "strong" | 1of10, as community-observed ranges | Creator-reported. The same page says YouTube publishes no official benchmarks. |
| Your viewed rate must reach 81.1 percent to be pushed | A claim that circulates | Satura says YouTube does not publish an 81.1 percent viewed-rate requirement, a public push threshold or a formula that guarantees distribution. |
| Average view duration should exceed 60, 50 and 40 percent of runtime for Shorts under 15, 15 to 30 and 30 to 60 seconds | Retensis | The vendor's own targets. The page adds that no platform publishes a cutoff. |
| An engaged view means 30 seconds watched | Socialinsider | Conflicts with YouTube, which defines it as a stay beyond the opening seconds, loops excluded. |
| YouTube counts a view at one second | Retensis | Not what YouTube's Help says: a Shorts view counts a start or replay with no minimum watch time. |
| Raw Shorts views jumped 20 to 30 percent overnight in March 2025 | 1of10 | No source given. Look at your own channel around 31 March 2025. |
YouTube's guidance points to your own history instead. Its audience retention page offers typical retention, which compares a video with your ten latest videos of similar length, and its discovery page says its system does not favor any format of Short and sets no minimum posting cadence.
Build your own baseline and a log
A baseline is the middle value of your last ten Shorts of one type, for each number. YouTube's typical retention compares a video with your ten latest videos of similar length; we borrow the number ten and narrow the group to Shorts of one type, which is our working rule and a looser match than YouTube's own. Divide each new Short's number by that middle value to get a ratio, and treat a ratio below 0.90 as clearly below usual. Small moves either side of 1.0 are noise.
Give no verdict until the Short is at least three days old and has at least 500 engaged views. Three days covers the one to two days YouTube says retention data takes to process. YouTube publishes no minimum sample for Shorts, so 500 is our working floor: below it, a handful of viewers can swing the averages. Log the row anyway and judge it next week.
A new channel will not have ten Shorts of one type. With three to nine, use all the Shorts of that type you have. Below three, compare against all your Shorts and mark the verdict as provisional in the notes. The log follows the same rule: its ratio cells stay blank until three earlier rows exist.
The weekly review log holds one row per Short, and you keep one copy per type. You type in the date, title, type, length, shown in feed, how many chose to view, engaged views, average view duration, average percentage viewed and subscribers gained. Three ratio columns fill themselves in: shown in feed, how many chose to view and average percentage viewed, each divided by the middle value of up to ten rows above it. A fourth computed column gives subscribers per 1,000 engaged views, an analyst-made comparison that is not a named YouTube metric; use it to rank your own Shorts, not to predict anything. The last two columns are for your verdict and your notes.
Keep Shorts of one type together. A tip with a screen recording, a customer question and a product demo each have their own usual, and YouTube makes the same point: audience behavior differs by format, so success looks different. If you are still choosing what to film, the YouTube Shorts strategy guide covers the plan, and the hook analyzer can help you draft an opening to test.
Worked example: one founder's routine-tip Shorts
The numbers in this example are invented to show the method and come from no real channel. Picture a founder who sells a face cleanser and posts a 30-second routine tip as a Short most weekdays. Her log holds ten earlier routine tips. Their stay rates (how many chose to view) were 55, 57, 58, 59, 60, 61, 62, 63, 64 and 66 percent, so the middle value is 60.5 percent, halfway between the fifth and sixth. Shown in feed for the same ten was 7,200, 7,900, 8,400, 8,800, 8,900, 9,100, 9,300, 9,700, 10,400 and 11,600, a middle value of 9,000, and average percentage viewed was 64, 66, 67, 68, 69, 70, 71, 72, 73 and 75 percent, a middle value of 69.5 percent.
Her newest tip, Short A, opened on a close-up of the bottle while she read out the ingredient list. Four days after posting, with about 4,400 engaged views, it was ready to judge, and the log showed this:
| Number | Short A | Middle of the last ten | Ratio |
|---|---|---|---|
| Shown in feed | 9,300 | 9,000 | 1.03 |
| How many chose to view | 47% | 60.5% | 0.78 |
| Average percentage viewed | 72% | 69.5% | 1.04 |
Reach holds at 1.03, so YouTube gave the Short a fair test. Stay is 0.78, clearly below the 0.90 line, and depth holds at 1.04. That is the row "Stay is low, depth holds" in the pattern table: remake the opening and leave the rest alone. She kept the script, the length and the topic and changed only the first second, which now shows the finished result (clean skin in daylight), with the bottle arriving in the third second. Remake B reached a 59 percent stay rate, a ratio of 0.98, above the line, and its other two ratios stayed close to 1.0. Had B also come in below 0.90, the idea would have been dropped.
One remake proves little on its own. What she gains is a dated note for this type of Short: opening on the result held viewers better than opening on the product. She also tags each row with its opening (result or product), so after a month she can compare every routine tip that opened one way with every tip that opened the other way, inside the same type.
Drawn from this article's worked example, with invented numbers to show the method. Original diagram; no real channel, benchmark or result is shown.
Read the retention graph last
The graph for one Short is the most detailed view in the report, and it helps only once the earlier numbers say where to look. YouTube's audience retention page describes four shapes. A flat line means viewers watched that stretch to the end. A gentle slope means interest fades over time. A spike marks a stretch that people watched again or shared. A dip marks the point where they abandoned or skipped.
Two limits are easy to miss. YouTube says a video should be at least 60 seconds long with at least 100 views for the highlighted key moments (intro, top moments, spikes and dips), and they show only when detected, so a 20-second Short with few views gives you the shape of the line and little else. And retention data takes a day or two to process, so read the graph after the first weekend, not the first hour. When you find a dip, watch the Short at that second and write one guess about the cause. A dip is a timestamp, not a diagnosis.
The four shapes follow YouTube's audience retention Help page, read November 7, 2026. Original diagram; not data from a real Short.
Run a fifteen-minute weekly review
- Wait until your newest Shorts are a few days old so retention has processed.
- Open Analytics, Content, Shorts, and set the range to the last 28 days. Note shown in feed, how many chose to view and subscribers.
- Open each of the last three or four Shorts and add one row to the log for its type: the six numbers, its length and its date.
- Skip any Short under three days old or under 500 engaged views. For the rest, read the three ratio columns and note which are below 0.90.
- Read the comments on the week's strongest and weakest Short, and copy any line that gives a reason in the viewer's own words into the notes column.
- Give the Short a verdict from the pattern table and enter it in the verdict column, with the one change for the next Short of that type and the date in the notes column.
- The week after, read the same number for the Short that carried the change, then keep or drop the change.
Change one thing at a time, as Satura also advises, and keep the audience promise the same, so the number you compare reflects the edit and not a different topic.
Plan next week's Shorts from the verdicts
The verdicts become next week's posting plan. Fill your slots in this order, which is our working rule: first one remake for each remake verdict, changing only the failing part; then repeats, each on the same premise with a new example; then any format whose reach was low, posted again on a fresh topic; and last, one new idea, so the log keeps testing something fresh. Drop verdicts free up slots. Keep the number of slots you can sustain; YouTube's discovery page says there is no minimum posting cadence.
For the founder in the worked example, who posts five routine tips a week, the plan after that review could read: Monday, Remake B with the result-first opening; Tuesday and Wednesday, two repeats of her strongest recent tip with new examples; Thursday, a tip whose reach was low, posted again on a different skin concern; Friday, one new idea, such as a customer question. Next week's review then reads the stay ratio on Monday's Short to decide whether the new opening stays. The short-form strategy guide shows where this sits in a wider plan.
Where Superdirector fits
Superdirector publishes this guide, and it does not replace YouTube Studio: your own Shorts report, with shown in feed and the retention graph, stays there. What it adds is the outside view. It tracks accounts, competitors, keywords and hashtags on Instagram, TikTok, YouTube and X, with their posts and metrics, so you can see which formats are taking off in your niche before you choose the next Short to test.
There are two plans. Starter, at $29 a month, works from Claude or ChatGPT for hands-on research, ideas, scripts and downloads. Growth, at $199 a month, runs a social account for you with AI: it makes carousels, X posts and video scripts in your brand, fills the month's calendar and posts carousels to Instagram and text posts to X. Finished AI video and full autopilot, which would include posting Shorts, are coming soon. See what runs today.