EVIDENCE 006
From talks with Youtube AI assistant
Sometimes reality forces even a robot to forgot whose side it was supposed to be on...
YouTube Studio’s AI assistant concluded that Shorts were feeding the algorithm the wrong data and recommended backing away from Shorts
After analyzing the channel, it concluded that Shorts were feeding the algorithm the wrong data and choking the long-form videos with irrelevant recommendations.
Transcript of the correspondence:
If you decide to contact TeamYouTube (X/Twitter) or support chat, use these facts:
1. “The channel has 0 Content ID complaints and 0 Community Guidelines violations.”
2. “Audience retention on the new long-form videos remains at 35–50% during the first minute, which is a strong result for the music niche.”
3. “Despite high retention and CTR above 4% on some videos, the number of organic impressions is not growing, while the traffic sources (Suggested) show abnormally low relevance.”
These are technical arguments indicating a problem with content classification in the recommendation system, rather than poor video quality.
My verdict: You are currently “feeding” the algorithm the wrong data through Shorts. Your long-form videos are technically ready for growth, but they are being “choked” by irrelevant recommendations.
The safest step now is to let the algorithm “cool off” from Shorts and try to change the geography through the Community Tab (an English-language poll, GIF) in order to attract a more relevant audience segment.