Data analysis from YouTube Analytics

Youtube Analytics provides a lot of information about the performance of the channel, including our audience and the achievements of individual videos, but the interpretation of this data is a common problem. Guides available on the Internet usually indicate a few basic metrics, such as the number of views or watch time, which we should observe, adopt as KPIs for our channel and strive to make them as high as possible at all costs.

However, we would like to point out that the analysis of this data does not boil down to just comparing a given metric in a group of films and stating that the film with a higher score is better than the others. In order to draw reasonable conclusions, in most cases it is necessary to compare a given metric with others that are available in YouTube statistics and only on this basis can we assess the real results of our video or the entire channel.

Below are some of the most commonly used metrics, along with some tips to consider when analyzing them.

Impressions/Views 

The number of views used to be a basic indicator and important information for the algorithm on the basis of which YouTube assessed whether a given video was willingly watched by viewers. We can imagine what room for abuse and manipulation was created by this solution -  it did not matter whether the viewer actually watched the material or just entered the player's page and left it immediately.

Is a video with 10,000 views and an average view time of 60 seconds as valuable as a video of the same length and topic with 10,000 views and an average of 5 minutes of views? Currently, a more relevant and authoritative metric is watch time, which, when combined with the number of views, gives much more reliable information about whether a given video is valuable when compared to other similar videos.

When analyzing the number of views, it is therefore worth taking into account the length of videos, their average watch time or average watch percentage, and the total watch time.

Watch time

Watch time is a very important metric and one of the criteria for applying to the YouTube Partner Program, where a channel is required to generate at least 4,000 hours of total watch time in the last year. Theoretically, this criterion can be met by creating a lot of videos with an average view percentage of 15%. But will this content be as valuable as a much smaller number of videos of similar length but with an average view percentage of around 50%? Of course not.

Remember that YouTube focuses on quality content and viewer engagement. A channel that creates content with a higher percentage of views in a group of similar videos (similar length and subject matter) will be better promoted and thus will grow faster.

Total watch time by itself tells us little. We suggest analyzing it in conjunction with video length and average watch percentage or average watch time.

Subscribers

Subscribers, i.e. the difference between the number of subscribers lost and gained in a given period. The more, the better, but does the number of subscribers alone tell us how effective a given video was in attracting them compared to other similar videos?

We suggest considering the number of acquired subscribers at least in the context of the number of views.

However, given that the number of views is made up of both subscribed and non-subscriber viewers, it is worth reaching for the filters in advanced statistics and comparing the number of subscribers with views that come only from unsubscribed viewers.

Thumbnail click-through rate

CTR, i.e. thumbnail click-through rate, is a metric that shows us what percentage of viewers opened our video after YouTube displayed the thumbnail of this video on the website. The success of the video and the entire channel largely depends on the attractiveness of the thumbnails and titles, and this is one of the most important factors when creating valuable content.

CTR is therefore a very important metric as it measures the effectiveness of the title and thumbnail.

Considering the entire "lifecycle" of a video on YouTube and the fact that subscribers are more likely to watch our videos than non-subscribers, it is natural that in the initial phase after the video is published, the click-through rate is higher (a lot of views from subscribers), and starts to decline after a while.

The click-through rate also depends on the topic of the videos and external factors, including the number and quality of our competitors' videos on YouTube. Therefore, it is difficult to determine which CTR is considered high and which is too low. YouTube once gave a range of 2 to 20% as a good ratio, so the range is quite large.

Therefore, we suggest striving to ensure that the click-through rate of subsequent published materials is higher than the average of our channel.

Unique viewers

Data on unique viewers is only available for the last 90 days. The higher the number of unique viewers, the more new subscribers we can get. In turn, juxtaposing this information with the number of views gives us valuable information about which videos viewers are most likely to return to.

Average watch time and average view percentage

Both metrics depend to a large extent on the length of the content we publish on the channel, so  we cannot precisely determine what percentage of views should be considered high and what percentage should be considered low.

We mentioned earlier that it's good to analyze this data in conjunction with the number of views and watch time, taking into account the length of the videos. This is because very long videos tend to have a lower average view percentage, while short content will have a higher average view rate because it's shorter.

By analyzing the results of videos on a given topic, we can select those that had the highest percentage of views and generated the most total viewing time, and in comparison with the average viewing time, we can determine what length of videos is most suitable for viewers.

To sum it up, when visiting YouTube Analytics, let's try to verify and investigate what the numbers are based on, how they are influenced by the length and subject of the videos, and whether the total statistics of the video group are really satisfactory in the context of the number of videos published. More is not always better.