Google-owned YouTube says that people can manage their video recommendations through the feedback tools the platform offers. But do YouTube’s user controls actually work? A new investigative report from Mozilla and co-author Jesse McCrosky, who also writes for DataEthics.eu, shows that they really don’t.
22,722 volunteers donated their data to the non-profit organisation Mozilla, who’s behind the non-intrusive browser Firefox. The data – including recommendations of more than 560 million YouTube videos – was scrutinized to determine how much control people actually have over the YouTube’s recommendation algorithm.
Here are the top findings:
- People don’t trust YouTube’s user controls. More than a third (39.3%) of the people surveyed felt YouTube’s user controls did not impact their recommendations at all. Said one participant: “Nothing changed. Sometimes I would report things as misleading and spam and the next day it was back in […] Even when you block certain sources they eventually return.”
- Data confirms people are right. Mozilla’s analysis of 560 million recommended videos revealed that using the “Do not recommend channel” button prevented only 43% of unwanted recommendations. Other controls had even less of an impact: The “Dislike” and “Not interested” buttons only prevented 12%.
According to the executive summary of the report, there are examples of lost control. One participant asked YouTube to stop recommending firearm videos — but was shortly after recommended more gun content. Another person asked YouTube to stop recommending cryptocurrency get-rich-quick videos, but then was recommended another crypto video.
Some of the recommendations of the report is that YouTube should design its feedback tools in a way that puts people in the driver’s seat, and that YouTube should provide researchers with access to better tools that allow them to assess the signals that impact YouTube’s algorithm.
If you are interested in signing a petition to force YouTube to their users more control, do go to Mozilla’s website.