n 2006 Netflix offered a $1 million prize for anyone who could improve its movie preference recommendations by 10%. Netflix, at the time, made most of its money sending DVDs in the mail to users’ homes

Mathematicians went wild. The competition was lauded by business pundits as an example of crowdsourcing genius. Because this was damned hard math, the project took years. And then in 2009, a team of mathematicians called “BellKor’s Pragmatic Chaos” actually cracked the code, achieved a 10% lift, and Netflix gave them the $1 million.
And then … Netflix never implemented the winning algorithm. Because personalization at that point no longer mattered.

Personalization has been such a buzzword for so many years. Netflix was one of the poster children for this. It's interesting to look at articles like this and understand they really don't utilize it like say an Amazon does.

In fact, this article is very critical of Amazon and I'd have to agree. Amazon has decent recommendations, but it seems to be a fairly basic market basket model that shows what others who bought similar items. That may be the best way to offer items to customers. Amazon has all the money in the world for R&D, in fact they flaunt how much money they put back into their business and if they are using this model, it must mean the personalization models of predicting other types of product must not bring in as much as the market basket.