External evaluators and recommendations

We’re always working to improve recommendations. Our engineering teams continually experiment and test improvements to search results and recommendations. We use these test results, along with user feedback and other signals, to update and improve recommendations over time.

YouTube also uses third party evaluators to provide feedback on YouTube’s search results, recommendations, and the relevance of certain videos. This feedback helps us evaluate, train, and improve our systems, as well as the quality of recommendations on YouTube. These evaluators are trained using the same public guidelines that Google uses for search results.

These evaluators are different from our Community Guideline reviewers, who focus on enforcing our content policies and removing videos that violate our Community Guidelines.

Was this helpful?
How can we improve it?