I've been following an interesting conversation at the ALA group on LinkedIn about a recent article in the Wall Street Journal, Your E-Book is Reading You. I suppose it was only a matter of time before we all found out that eReaders can and do collect data about us.
It shouldn't be a surprise. Advertising and merchandising has always been dependent on an understanding of the customer. Being able to find out so much about what readers do and do not like about their books is a treasure trove to publishers.
One problem I see is that it's early days for eReaders. Not everyone uses an eReader, not everything is available on eReaders. So the data set publishers are using reflects the reading tastes of only a percentage of all the readers and potential readers out there.
Meanwhile, the data they're collecting: how long it takes a reader to finish a book, where they stopped, what they highlighted, doesn't tell the whole story. Unless you know why you don't have the whole picture. Maybe I stopped reading on page 53 because the kids got sick, maybe it took 2 weeks to finish the book because that was when the dog buried the reader in the backyard, maybe I let my 5 year old highlight that passage just to keep him occupied at the doctor's office.
As one comment noted, we have opened Pandora's Box and it's too late to go back now. If publishers and authors can use the data to create books that reach more people (and sell more copies) that's a good thing. If they end up stifling creativity to reach the least common denominator of readers that would be sad.