QPrime boost: a YouTube boost named after a shopping query classifier
The leak describes QPrime as a query classifier that asserts shopping categories. A YouTube twiddler carries its name and annotates each result. That it reads QPrime's output rests on the name; what it boosts is not documented.
01What the schema shows
The schema object is quality_youtube.QPrimeBoostTwiddlerResultAnnotation: an annotation attached to each result. "Boost" in the name suggests the twiddler lifts some results rather than filtering them out. Like every twiddler, it only recommends; the framework in Superroot, the layer that collects and merges results before they are shown, makes the final call.
RESONEO connects the twiddler to QPrime, which it reads in the leak as a query classifier for shopping categories, and says the boost depends on that classifier.
02What the leak says about QPrime
The 2024 leak of Google's API documentation has a model named RepositoryWebrefCategoryAnnotationShoppingSignals. Its description reads "The QPrime query classifier signals." One of its fields, isShoppingAnnotation, is described as "Whether QPrime asserts that particular category."
So QPrime classifies queries and can assert a shopping category. The model says nothing about YouTube and nothing about a boost. The step from this classifier to the YouTube twiddler is made by the shared name, not by any text in the leak.
03E-E-A-T and the system map
If the name means what it says, QPrime boost lifts certain YouTube results when a query has been classified as shopping: product reviews, unboxings or demos for a product search, for example. Which videos qualify, how strong the lift is, and whether it acts in YouTube search or in Google's video results is not known.
This is an intent rule for a commercial vertical, not a quality judgment. It has no E-E-A-T bearing and sits on the twiddler layer of our map.