A bit more than a decade ago I met two refugees from Google and Bing who were jaded about the failure of personalization projects at those companies. Both said that they these companies were not able to deploy deeply personalized search because they weren't able to evaluate deeply personalized search. A survey of the industry I did later showed that very few, if any, enterprise search vendors had a serious evaluation effort because such an effort requires a stupdendous amount of data and the vendors were motivated by the needs of people buying the software (who want 300+ "integrations") and not the experience of people using the software.

Bing and Microsoft, on the other hand, had enough data that they really could tune up search for everyone but not enough to tune up search for individuals, or to prove that tuned up search for individuals was really better for those individuals. So they didn't do it.

Googlers around this time would deny this and tell me "Google Search is highly personalized" but Google keeps Googlers in the dark about how search ranking works and almost all of them know less than middle-of-the-pack SEOs. At that time Google and Bing used cheap personalization algorithms that were so certain that they didn't need personalization, such as geolocalization ("thai restaurant" is "thai restaurant in your area") and reranking up results that you frequently click on (boy does it gaslight SEOs when they see their own results moving around!)

Today they can personalize because they just don't care if the results suck.