YouTube’s source code has just been leaked, completely revealing this platform measure used to control YouTubers

Not only that, but it also shows YouTube’s bias towards business brands, celebrities, rather than independent YouTubers who make the platform successful.

Recently, YouTube’s codebase, including the platform’s internal metrics, including a video “restricted display” label, has been leaked. These three internal metrics show the scores and label labels used to determine whether a video is restricted from displaying on the platform.

After this incident, YouTube had to secretly hide its code to avoid prying eyes.

Previously, these leaked measurements appeared from October 8 to October 29, before being silently hidden and hidden by YouTube. Earlier, YouTubers like Bowblax, Nicholars DeOrio, Optimus and Josh Pescatore had the opportunity to dig into this code and analyze their impact on more than 200 different channels.

Their findings not only show that YouTube is using its proprietary internal metrics to control YouTubers, but also giving preference to business brands, Hollywood stars, and other Channels are called “authentic sources” by YouTube.

1. The measure of P-Score score

The first metric leaked through the codebase above is YouTube’s P-Score priority score (Preference Score). P-Score is an internal measure to promote Google Preferred – a network of brand-safe advertising channels, accounting for about 5% of the content on YouTube.

According to YouTube, the P-Score is designed to push “the most engaging and engaging brand-related content on YouTube” to Google Preferred advertisers. Before being discovered in code lines, the P-Score score was supposed to be an internal measure hidden.

The leaked code shows that the P-Score score applies to all YouTube videos, whether or not it is included in the Google Preferred program. Basically, YouTube uses this score to evaluate the ability of each video to offer to potential advertisers.

These P-Score scores are exclusive, meaning that YouTubers will not know how to increase them. The only information that YouTubers know about the calculation of this P-Score score is in a video posted in April 2019 showing 5 common signals to improve this score:

1. Popularity: (Popularity is assessed based on watch time).

2. Passion level (Passion: based on the level of attracting viewers to the channel).

3. Safety level (Protection: based on whether the content is safe for the brand or not).

4. Platform (Platform: which featured content is viewed on the big screen).

5. Production capability (Production: including content that uses “skilled camera techniques and cinematography techniques”.

After the P-Score score was discovered in the above codebase, Bowblax, Nicholars DeOrio, Optimus and Josh Pescatore said their findings show that the P-Score score is a measure based on region and region. YouTube channels will usually have a higher P-Score score in their home countries.

2. The measure of label evaluation of content labels

Another measure of leak this time is the content label rating according to a 6-level rating system. Research shows that this content label rating system affects the ability to monetize and distribute content on YouTube according to the following criteria:

1. Y: Content is especially safe for everyone who watches YouTube.

2. G: Especially family friendly. Achieving this criterion is likely to have a high CPM (cost per 1000 impressions – cost per 1,000 impressions) and may be included in the Protection category and have an impact on P-Score scores.

3. PG: Family-friendly in general. Most likely have excellent CPM levels and can be included in the Protection category and have an impact on P-Score scores.

4. Teen: occasionally suitable for the family. Most likely to have a good CPM with some restrictions on advertising.

5. Mature: rarely family friendly. Most likely to have a low CPM if monetization is enabled.

6. X: Default age limit. Never make money.

3. The limited measure of display

The third measure leaked in this code shows the possibility of limited display and the limited number of evaluation scores displayed.

According to YouTuber’s research, there are two types of display restrictions:

– display restrictions for brand safety: limit advertising on the video.

– Limited display for special videos: algorithm limited the ability to distribute videos on the platform.

4. Impact of P-Score

Besides discovering internal leaks in the code, YouTuber Bowblax, Nicholars DeOrio, Optimus and Josh Pescatore also compiled P-Score scores for over 200 different channels.

The results reflect most of what we see on YouTube every day – business brands, Hollywood stars, and “authentic sources” dominate most of the content displayed, while independent YouTubers are ranked lower on this list, even though they are the people who brought this site to where it is today. Only one YouTuber, Linus Tech Tips, made the list of P-Score points in the top 10.

Source : GenK