For all the bigger companies I've worked with that have infinite war chests and rely heavily on user data and analytics, the question is always why shouldn't a specific metric be tracked, not why should it. Even if they have absolutely no reason to track how fast you're reading, it's safe to say they're tracking it anyway. At a certain point you stop reading privacy policies or listening to what a company says it's doing or not. You even stop looking at code or inspecting network traffic. You simply assume you're being tracked in every technical way it's possible for "them" to do so, and you either accept it or change it.

It could also be a useful proxy for density/interestingness of the material (when adjusted for your typical reading speed). Fed into a recommender system it might produce better outputs and more purchases.

It could also be sold to publishers as part of a tool to determine the above prior to publishing; what if books were optimized for your attention?

Pleasant or not, there are plenty of uses for the data.

Probably amazon

They can also see the things that you highlight, the things that you reread. This could be used to profile the reader and personalize ads

not only this, but being forced to buy an ebook through their portal, they'll know everything you're reading. Whereas you could read pdfs anonymously if you wanted

The Kindle app tracks how fast you're reading. If nothing else it uses it to improve it's estimate of how long until you finish reading the chapter, which it tells you.

The probably track it to make sure you are not gaming the system. Kindle Unlimited pays out based on pages read.

https://kdp.amazon.com/en_US/help/topic/G201541130

> A customer can read your eBook as many times as they like, but we will only pay you for the number of pages read the first time the customer reads them.