ContentMate
    YouTube

    YouTube metrics explained

    YouTube metrics explained: views, engaged views, watch time, CTR, subscribers, retention, playlists, revenue, and CPM with formulas and examples

    What counts as an engaged view? Why do views differ from thumbnail impressions? Which numbers help you evaluate a Short? This reference explains the individual YouTube metrics in ContentMate, their units, calculation examples, and limitations.

    Names follow the labels in Analytics. Not every report provides every metric: availability depends on the content type, channel, monetization, permissions, and data. For opening reports and selecting dates, see Understand YouTube analytics.

    Before comparing: period, unit, and data freshness

    Period metrics describe activity during the selected dates. All media instead shows a video's latest imported lifetime counters. The subscriber count is a snapshot. For example, a video with 10,000 lifetime views might have received only 600 views during the selected month.

    YouTube assigns daily data to America/Los_Angeles. Dates that have not finished processing can be missing. ContentMate imports in the background; opening a chart does not make a live request to YouTube. A dash means unavailable or not safely calculable; an explicitly reported 0 means zero.

    Counts can be added across matching dates within the same report. Percentages and averages require their own denominators. Simply averaging the displayed daily percentages is often wrong. Additional daily reports may cover only dates already imported; check the completeness notice.

    Views and watch time

    Video views

    Unit: count. YouTube counts playback starts. The changed start-based counting applies to Shorts from 2025; in August 2026 YouTube also aligned counting for long-form videos and live streams. A start can therefore count even when little further viewing occurs. Repeat playbacks are not additional unique people.

    Evaluate views alongside engaged views and watch time. A jump compared with older periods may also reflect a counting change. YouTube: changes to view counting.

    Engaged views

    Unit: count. Under YouTube's current definition, playbacks that continue past the first frame, or start through an explicit click or tap. This helps distinguish initial starts from more deliberate viewing. It does not count unique viewers.

    If a Short has 2,000 views and 900 engaged views, these are two ways of counting activity on the same content. The 900 are neither additional views nor necessarily 900 different people. Their ratio is also not automatically the separate “stayed to watch” metric in YouTube Studio.

    YouTube Premium views

    Unit: count. Views by viewers with YouTube Premium. These describe part of the video views and should not be added on top of them. One hundred Premium views do not mean 100 new Premium memberships or specify the resulting revenue.

    Watch time (minutes)

    Unit: minutes. Total viewing time reported by YouTube. A value of 120 means two hours, not 120 viewers. More watch time can come from more playbacks, longer videos, or longer viewing. Also check average duration and average percentage viewed.

    Premium watch time (minutes)

    Unit: minutes. Viewing time from YouTube Premium members. It is part of the total watch time. Together with Premium views it helps explain viewing activity, but there is no fixed price per minute that converts it into Premium revenue.

    Average view duration (min:sec)

    Unit: minutes:seconds per included playback. YouTube's average viewing duration; 0:45 means 45 seconds. For Shorts, the average uses engaged views and their corresponding watch time. Do not automatically divide all watch time by every playback start.

    In a simplified example, 600 matching viewing seconds across 20 included playbacks produce an average of 30 seconds. Prefer comparisons between videos of similar length and format. ContentMate uses the relevant imported report value; daily averages are not added together.

    Average percentage viewed

    Unit: percent. How much of a video was watched, on average, during an included playback. For one 60-second video, an average of 30 watched seconds represents 50% in a simplified example. This does not mean that 50% of people completed the video.

    Across videos of different lengths, “total viewing duration ÷ average video length” is not a reliable replacement calculation. For Shorts, engaged views and their corresponding watch time are again the relevant basis. YouTube: viewing and content metrics.

    Reactions and shares

    Likes

    Unit: count. Positive video ratings recorded by YouTube in the report. A like is a rating, not an additional view. Likes within a selected period can differ from a video's current lifetime counter. YouTube can subsequently correct invalid ratings.

    Dislikes

    Unit: count. Recorded negative ratings on your videos. Read them alongside likes, comments, and audience retention to understand the response to content. A few dislikes on very few views do not establish a trend; this metric is separate from the publicly visible like counter.

    Comments

    Unit: comment actions in the report. How often people commented. One person can leave multiple comments, so the analytics value is neither the number of distinct authors nor the number of open inbox conversations. Deletions, moderation, and processing delays can complicate comparisons with the visible comment list.

    Shares

    Unit: count. Sharing actions through YouTube's sharing feature. The Sharing services report groups these by the service used. The count does not say how many recipients received or watched the link. Manual sharing outside the tracked feature is not a fully measurable additional share.

    These values are not converted into a universal cross-platform engagement rate: reactions, people, and views measure different things. YouTube's engagement definitions explain the individual events.

    Subscribers, audience, and publications

    Subscriber count

    Unit: count from the latest imported snapshot. The channel count supplied by YouTube can be rounded. It describes channel size as of the data timestamp. Comparing two rounded snapshots does not give an exact daily subscriber change.

    Subscribers gained

    Unit: count within the period. New channel subscriptions. Channel reports also include subscriptions outside video watch pages. A report for one video counts the subscriptions attributed there, so adding video rows does not necessarily produce the channel's total gain.

    Subscribers lost

    Unit: count within the period. Ended channel subscriptions. This is the number of losses even when shown as a positive number. Net change = subscribers gained − subscribers lost: 35 gains and 12 losses give +23. Paid channel memberships are a different feature.

    Viewer share

    Unit: percent. Share of the logged-in audience represented in a demographic report, for example a particular age and gender combination. 25% refers to the reported demographic population, not automatically all of the channel's viewers.

    YouTube may withhold small groups. Percentages from different days must not be added or averaged without suitable weighting. “Country,” “subscription status,” and “device” are breakdowns: views in the “Mobile” row are still views, not unique people. YouTube: demographic dimensions.

    Posts or published videos

    Unit: number of imported videos. ContentMate counts videos with a publication date inside the selected period. This measures publication volume. It does not include videos not yet imported, and does not measure views or YouTube community text posts.

    Thumbnail impressions and click-through rate

    Thumbnail impressions

    Unit: count. Recorded displays of video thumbnails on supported YouTube surfaces. YouTube's definition requires the thumbnail to be visible for more than one second with at least half of it on screen. External embeds and various other surfaces are excluded.

    One person can generate several impressions, so this is not unique reach. Equally, not every video view must originate from a counted thumbnail impression. YouTube: where impressions are counted.

    Thumbnail click-through rate

    Unit: percent. Share of counted thumbnail impressions that lead to a click to watch. CTR = corresponding clicks ÷ thumbnail impressions × 100. Four hundred matching clicks from 10,000 impressions give 4%. Do not use all video views from all sources as the numerator.

    Several days require impression weighting: 100 impressions at 10% and 900 at 2% give 2.8%, not 6%. ContentMate checks imported coverage for the corresponding period metric; missing inputs are not replaced with zeros.

    Thumbnail click-through rate (YouTube original value)

    This label appears in additional daily reports. There, the number keeps YouTube's supplied representation instead of adopting the normal percentage card's formatting. Do not compare it to the formatted Thumbnail click-through rate solely by the visible number, and do not add the rows. Its meaning is the same; the display differs.

    Audience retention and video segments

    These metrics describe positions inside a video. The 50% position in a 60-second video is at 30 seconds; it is not a calendar date. Open the video's details to investigate particular scenes.

    Audience retention

    Unit: percent at a video position. How often a segment was watched relative to the video views in the report. With 100 views and 60 plays of that segment, the value there is 60%. Watching the same section again can raise the value above 100%.

    A decline can indicate where playbacks end or sections are skipped. A peak can arise from repeat viewing. The curve is not a list of identifiable viewers.

    Relative audience retention (0–1)

    Unit: comparative score from 0 to 1. Retention at a video position compared with YouTube videos of similar length. 0.5 is the midpoint: roughly half retain viewers better there and half worse. A value of 0.8 does not mean “80% are still watching.” Use the normal audience retention metric for that question.

    Playbacks started per segment

    Unit: count. How often a segment was the first section watched in a playback. Entering partway through a video or resuming it can therefore produce a start in the middle. Not every counted start happens at second zero.

    Playbacks stopped per segment

    Unit: count. How often a segment was the last section watched in a playback. This can be an early exit or the normal end of the video. High values at the ending therefore mean something different from an unusual number of exits just after the opening.

    Segment plays

    Unit: count. How often a section was watched, including repeat viewing. Someone watching the same section twice can produce two segment views. Adding all sections therefore does not give the number of complete video views. YouTube: audience retention metrics.

    Concurrent viewers in live streams

    Average concurrent viewers

    Unit: average number watching at the same time. Describes the simultaneous audience for the stream or displayed stream segment. A value of 40 does not mean that only 40 people watched in total. Different people arriving and leaving can produce the same average.

    Peak concurrent viewers

    Unit: count at the peak. Largest simultaneous audience in the stream or segment. Average and peak answer different questions: an average of 40 and a peak of 90 indicate a period of stronger attendance. Peaks from individual sections must not be added to form a total audience. YouTube: live-stream metrics.

    Videos in playlists and playlist use

    Adding a video to a playlist and saving a whole playlist are different actions. Playlist reports also distinguish viewing within a playlist from general views of your own videos that belong to it. Do not add these groups together.

    Added to playlists

    Unit: video-addition actions. Includes your playlists, other channels' playlists, and Watch Later. Automatic additions to upload lists or watch history are excluded. Adding, removing, and adding the video again produces two additions: actions are counted, not distinct people.

    Removed from playlists

    Unit: video-removal actions. Also includes removal from Watch Later. Removing a video again after adding it back counts again. This is neither the number of deleted videos nor the number of deleted playlists.

    Playlist views

    Unit: video views within the particular playlist. Videos owned by other channels can also contribute to playlist viewing. Watching a contained video outside this playlist does not produce a playlist view for this list. Watching three videos in succession can produce three playlist views.

    Playlist watch time (minutes)

    Unit: minutes within the playlist context. Total viewing time while watching this playlist, potentially including other channels' videos. Time spent watching the same video outside this playlist is excluded. This helps evaluate actual use of the list.

    Average playlist view duration (min:sec)

    Unit: average duration per video view within the playlist. Applies to individual videos watched, not the complete playlist session. 2:00 can mean that the included video views last two minutes on average.

    Playlist starts

    Unit: count. How often viewers started playing a playlist. YouTube reports this metric for playlist use on the web. One start can lead to multiple video views. Comparisons with playlist views across platforms therefore need the same coverage.

    Average views per playlist start

    Unit: average video views per start. Indicates how many videos are watched after a playlist start, on average; this metric covers web use. 2.5 means two and a half video views per start on average. Do not divide views from every device by web-only starts.

    Average time in playlist (minutes)

    Unit: average minutes after a playlist start on the web. Describes the playlist session as a whole. This differs from the average duration of an individual video view. Three videos watched for two minutes each could together make six minutes of playlist use.

    Playlist saves (net)

    Unit: net change within the period. Whole-playlist saves minus removed saves. Twelve saves and five removals give +7. This is not the lifetime number of people who have ever saved the list; the value can be negative.

    Playlist saves added

    Unit: save actions in the additional daily report. Every save counts, including saving again after a removal. Unlike the metric explicitly labelled net, this value has not already had removals deducted.

    Playlist saves removed

    Unit: removal actions in the additional daily report. How often users removed the whole playlist from their saved lists. On matching data, “playlist saves added minus playlist saves removed” produces the net change. YouTube: playlist metrics and daily reports.

    Cards and card teasers

    A card is an interactive reference inside a video; its teaser initially draws attention to it. Each has its own impressions and clicks. These definitions follow YouTube's card metrics.

    Card impressions

    Unit: count. Cards displayed in the opened card panel. YouTube counts each displayed card separately. Showing three cards in one panel can produce three impressions. This is therefore not a count of video views.

    Card clicks

    Unit: count. Clicks on the cards themselves. They indicate how often someone selected the linked content, without guaranteeing that it was subsequently watched in full or that a desired action was completed.

    Card click-through rate

    Unit: percent. Card clicks ÷ card impressions × 100. Fifteen clicks from 300 impressions give 5%. The denominator is card impressions, not all video views or teaser impressions.

    Card teaser impressions

    Unit: count. How often prompts for cards were displayed. Several teasers can appear during one video view. This measures opportunities to notice a teaser, not openings of the card itself.

    Card teaser clicks

    Unit: count. Clicks on those prompts. YouTube attributes clicks on the card icon to the last teaser shown. A teaser click and a subsequent card click describe different steps and are not merged into one click-through rate.

    Card teaser click-through rate

    Unit: percent. Teaser clicks ÷ teaser impressions × 100. Twenty clicks from 1,000 prompts give 2%. A strong teaser rate with a weak card rate may be a reason to examine the card's wording and destination.

    End screen elements

    Additional daily reports distinguish individual elements near the end of a video, such as a video recommendation or channel reference. They can be analysed only when YouTube supplies the corresponding data. YouTube: end screen metrics.

    End screen element impressions

    Unit: displayed elements. Each visible element is counted separately. Two elements in one reached end screen can generate two impressions. This is therefore not a count of people who watched the video to its end.

    End screen element clicks

    Unit: count. Clicks on these elements. They help show whether viewers follow an offered next step. Clicking a channel reference, however, is not automatically a new subscription.

    End screen element click-through rate

    Unit: percent. Element clicks ÷ element impressions × 100. Ten clicks from 500 impressions represent 2%. Compare similar element types, positions, and audiences where possible; video views are not the denominator.

    Historical annotations

    Annotations are an older YouTube format. These fields support historical reporting and do not promise a currently available publishing feature. Missing values for new videos are consequently not evidence that an import failed.

    Annotation impressions

    Unit: count. All historical annotation displays reported by YouTube, whether clickable or closable. Several annotations in one video view can produce multiple impressions.

    Clickable annotation impressions

    Unit: count. Displays where an annotation could be clicked. This subset is the denominator for annotation click-through rate. It does not necessarily equal all annotation impressions.

    Closable annotation impressions

    Unit: count. Displays where viewers could close the annotation. This measures the opportunity to close it; actual closing actions appear under Annotations closed.

    Annotation clicks

    Unit: count. Actual clicks on historical annotations. Do not merge these with clicks on modern cards or end screen elements: they are different formats with different impressions.

    Annotations closed

    Unit: count. How often viewers closed a historical annotation. The action may indicate an intrusive overlay, but does not by itself explain an exit from the whole video.

    Annotation click-through rate

    Unit: percent. Annotation clicks ÷ clickable annotation impressions × 100. Eight clicks from 200 clickable impressions give 4%. Non-clickable impressions do not belong in this denominator.

    Annotation close rate

    Unit: percent. The proportion of closed annotations under YouTube's supplied definition. The current English metric reference relates this to all annotation impressions. Other documentation passages may differ; do not independently replace the denominator with closable impressions. Use the matching imported report. YouTube: historical annotation metrics.

    Revenue and advertising

    ContentMate displays available monetary values in US dollars (USD). Revenue reports require an eligible, monetized channel and the relevant Google permission. Figures are estimates and may later change. They are not confirmed payout or tax figures. Here, “gross” and “net” describe YouTube's revenue allocation, not your personal VAT treatment.

    Estimated revenue

    Unit: USD within the period. Estimated net revenue attributed to you from advertising and non-advertising sources included in the report. It can include advertising and Premium revenue, among other sources. Do not add those components to the total again.

    Estimated ad revenue

    Unit: USD within the period. Estimated net revenue from Google-sold advertising. Its scope is narrower than total estimated revenue. Income from sources such as Premium does not belong to this advertising metric.

    Estimated Premium revenue

    Unit: USD within the period. Estimated revenue from YouTube Premium subscriptions attributed to the channel and report. It is not the total subscription price paid by viewers or a fixed payment per Premium view.

    Gross ad revenue

    Unit: USD within the period. Estimated advertising revenue before your ownership and revenue shares are taken into account under YouTube's definition. It is not your payout. Its coverage also differs from net revenue; the difference is not a reliable flat-fee calculation.

    Ad impressions

    Unit: verified ad displays. One video playback can include multiple ads. This metric is entirely separate from thumbnail impressions. For example, 100 playbacks with two ads each can generate 200 ad impressions.

    Monetized playbacks

    Unit: playbacks with at least one ad impression. Several ads during the same playback still count as one monetized playback. YouTube can also count a case in which a pre-roll ad is shown but viewing ends before the video itself starts.

    CPM

    Unit: USD per 1,000 ad impressions. Estimated gross advertising revenue on an impression basis. With matching data: gross ad revenue ÷ ad impressions × 1,000. Twelve USD from 3,000 ad impressions give a 4 USD CPM. This is neither your net revenue per 1,000 video views nor a payout promise.

    Playback-based CPM

    Unit: USD per 1,000 monetized playbacks. The same concept based on playbacks with ads: gross ad revenue ÷ monetized playbacks × 1,000. If the 12 USD came from 2,000 such playbacks, this gives 6 USD. Multiple ads per playback explain the difference from the 4 USD CPM in the example.

    Values must come from the same period and report filters. RPM, which relates to the creator's revenue, is a different metric and is not produced here from CPM or views as an additional YouTube metric. YouTube: revenue metrics and CPM and monetized playbacks.

    Membership cancellation survey responses

    Unit: completed cancellation surveys. The report counts responses from users who ended a paid channel membership and groups them by the stated reason. It does not count ordinary channel unsubscriptions.

    Five responses citing “too expensive” mean five reported survey responses, not necessarily every cancellation for that reason. Not everyone who cancels fills in the survey. A complete cancellation rate cannot be calculated without further denominators. YouTube: membership cancellations.

    Avoid common misinterpretations

    ObservationHow to interpret it
    Views rise while watch time stays the sameMore starts do not automatically mean longer viewing. Check engaged views and retention too.
    Click-through rate falls while impressions riseA different audience or source can affect the rate. Compare equivalent reports.
    Audience retention exceeds 100%Repeat viewing of a section can cause this.
    Video subscriber gains do not add up to channel gainsPeople can subscribe outside video watch pages.
    Daily reports show the same activity by device and sourceThese are different views of the same activity and must not be added.
    A metric is missing for a custom rangeRequired imported values or a sound calculation basis are missing; a dash is not zero.
    A Premium, playlist, or revenue report is emptyThe channel and period must have matching data. A report's presence does not guarantee values.

    Definitions checked on 10 October 2026 against ContentMate's displayed metrics and the official YouTube Analytics definitions, daily reporting definitions, and YouTube reporting changes. Google developer documentation is licensed under CC BY 4.0; explanations and examples here have been newly written for ContentMate.

    Was this article helpful?