YouTube Metrics: What to Look at After Publishing a Video

The YouTube metrics worth looking at after publishing a video are few: impressions and click-through rate, which tell you whether the title and thumbnail attract, and watch time, average view duration and retention, which tell you whether the video hooks people. Read them with enough data and by comparing similar videos on your own channel.

Where to see them: YouTube Studio Analytics

The figures for each video are in YouTube Studio: in the left-hand menu you select “Content”, hover over the video you are interested in and select “Analytics”. There are several tabs; for a newly published video these matter most:

• Overview: a summary of the video's performance, where you can also check the key moments for audience retention.
• Reach: how the audience discovers your channel. Its key metrics card shows impressions, impressions click-through rate, views and unique viewers.
• Engagement: how much time the audience spends watching your videos, with watch time and average view duration.
• Revenue: if you are part of the YouTube Partner Program, tracking of what you earn.

The basic metrics, with YouTube's definition

Metric	What it measures, according to YouTube	What it tells you
Thumbnail impressions	The number of times your videos' thumbnails were shown to users on YouTube. To count, at least 50% of the thumbnail must be visible and stay on screen for at least one second.	How many people had the chance to choose your video.
Impressions click-through rate (CTR)	How often users played a video after seeing the thumbnail.	Whether the title and thumbnail convince those who see them.
Views	The actual number of times your content has been watched. They are counted the moment playback starts.	Reach, with no indication of whether people liked it.
Watch time	The time users spent watching your video.	The total attention it received.
Average view duration	The average minutes of viewing from users who stayed to watch the content.	How long, on average, people who stay to watch keep watching.
Average percentage viewed	According to the YouTube Analytics documentation, the average percentage of a video that was watched during playback.	The same, in proportion to the video's length.
Unique viewers	The estimated number of people who have watched your content in the selected period.	How many different people are behind the views.

Two caveats from YouTube's help that avoid nasty surprises: click-through rate is based on views that come from counted impressions, so dividing views by impressions yourself may not give the same figure; and if a large volume of traffic comes from outside YouTube, you may have more views than impressions.

Audience retention: where people leave

The “Audience retention” report is only available at video level, in the Overview or Engagement tabs. Its key moments report shows how well the different parts of the video managed to hold users' interest, and its data takes 1 to 2 days to process. This is how to read the curve, according to YouTube:

• Flat stretch: users watch that part from start to finish.
• Gradual decline: they lose interest over time. As a general rule, YouTube videos experience some decline during playback.
• Spikes: more users watching, rewatching or sharing a particular part. Be careful, because a spike can also mean the content is unclear and the audience had to rewatch the section.
• Dips: users stop watching the video or skip to other parts.

Look at the introduction too: the percentage of the audience still watching your videos after the first 30 seconds. A high percentage can mean that those seconds match the expectations created by the thumbnail and title. To improve it, YouTube suggests changing the thumbnail and title so that they better reflect the content, or changing the first 30 seconds and trying different styles. For key moments to be highlighted, the video must be at least 60 seconds long and have at least 100 views.

When to look and how not to fool yourself

The figures do not all arrive at once. In the channel's Content tab, the summary of impressions, views and click-through rate is available a few hours after the video is published; retention takes 1 to 2 days. And if you read them from another tool through YouTube Analytics, the developer documentation states that processing generally introduces a latency of 48 to 72 hours.

On click-through rate, YouTube's help gives several pieces of advice:

• Do not make decisions without enough data – examine it after getting a considerable number of impressions and avoid checking it right after uploading the video.
• Do not react to small variations: it is normal for click-through rate to show small variations.
• Be wary of testing several titles or thumbnails on the same video: fluctuations may be due to traffic sources and not to the title or thumbnail.
• As a reference, for half of YouTube channels and videos the impressions click-through rate is between 2% and 10%. The range may be wider for new videos or channels, or for content with fewer than 100 views.

And compare carefully. Creators compare the videos on their channel with each other, but YouTube's systems rank each video against everything else a viewer could watch. In addition, it is normal for videos with fewer impressions and views to have a higher click-through rate and average duration, because they have been watched by a smaller, more loyal audience. Compare videos on your channel of a similar age and do not draw conclusions from a single figure. If what you want is to be found in search, start with title and description SEO.

Cost per thousand views: what it cost versus what it delivers

There are three “per thousand” figures that sound alike and have nothing to do with each other:

• CPM: the cost per thousand ad impressions before YouTube's revenue share. It is a metric for advertisers.
• RPM: your total revenue per thousand views after YouTube's revenue share. It is a metric for creators and you see it in the Revenue tab if you are part of the YouTube Partner Program.
• Production cost per thousand views: not a YouTube metric. It is what it cost you to produce the video divided by its views and multiplied by a thousand.

The last one tells you whether a video repays in attention what it cost: with similar views, the one that cost twice as much has twice the cost per thousand views. How to record what each video costs to produce is covered in how much it costs to make a video with AI.

How the Estudio de Canales tracks results

The Estudio de Canales does not publish to your channel: you upload the package in YouTube Studio and then paste the video's link into the studio.

• Linking. You can use the video's link (including youtu.be, Shorts or YouTube Studio links) or its ID. If the channel is linked, the studio checks with YouTube that the video belongs to that channel: if it belongs to another, it is not linked, and if it cannot be checked, it is linked and the studio says so.
• Figures. With your Google account connected, the studio reads the video's public figures (views, likes and comments) and, if the account is the channel's own, its daily evolution in YouTube Analytics, which arrives 48 to 72 hours late. What it cannot read is not filled in with zeros: it says why there is no data.
• History. If the video has been checked against YouTube, snapshots are scheduled at 1, 2, 3, 7, 14, 21, 30, 45, 60 and 90 days. Each one stores, per day, views, minutes watched, average duration, average percentage viewed, likes, comments, shares and subscribers gained and lost. The latest days are provisional and are rewritten in the next snapshot.
• Windows. The first 24 and 48 hours, the first 7 and 30 days, and the total. Each window says whether it is complete, and an incomplete one is not compared with others.
• What it does not show. Click-through rate, impressions and the retention curve do not appear in the studio: those are viewed in YouTube Studio.
• Cost per thousand views. It sets what the video cost to produce, according to the estimated cost of each operation, against what it is delivering.
• Channel learnings. It compares videos on your own channel without inventing causes: only with stored YouTube Analytics data and with the first 7 complete days of each video. A comparison by series, length or format needs at least 3 videos on each side, and each observation carries its limitations, starting with the fact that a comparison does not explain the cause. No AI, no “viral score” and no predictions.
• Back to ideas. From a video you can create a new idea for the channel; you write it yourself and it is noted which video it comes from.

Sources consulted

Checked on 9 October 2026. We link to the original source so that you can consult the current text.

• Empezar a usar Estadísticas de YouTube. Google · Ayuda de YouTube.
• Consultar las impresiones de miniaturas y el tiempo de visualización de YouTube. Google · Ayuda de YouTube.
• Medir los momentos clave para comprobar la retención de la audiencia. Google · Ayuda de YouTube.
• Preguntas frecuentes sobre las impresiones y el porcentaje de clics. Google · Ayuda de YouTube.
• Entender las métricas de ingresos publicitarios. Google · Ayuda de YouTube.
• YouTube Analytics API: métricas. Google for Developers.
• YouTube Analytics API: modelo de datos (actualidad y latencia de los datos). Google for Developers.