This is a very useful contribution, but something about it has been making me cringe, and I’ve trying to put my finger on it. Thanks to Bobbi Newman on Twitter, I think I’ve figured it out.

It’s wrong to divide this data into patron data and library data, because it’s all relationship data, and you can’t divide the relationship into the library’s part and the patron’s part. Instead, it’s better to characterize the uses made of the data and how those uses can strengthen or damage a relationship. It’s not the email address, it’s how the library uses the email address. Some uses can be opt-in, some opt-out, some required, others never done. In our big-data world, even checkout counts can be used to breach user privacy. (“Look how many times 50 shades is checked out on Jewish Holidays!”).