"Customers hated it from the bottom of their hearts," Dr Vogels says. "They hated the fact that there was somebody else just like them."
Amazon has also learned to tone down the veracity of several recommendation algorithms lest it risked 'creeping out' its customers. Such capability harkens back to the stories of US retailer, Target, using data to determine that its shoppers were pregnant.
"I'm pretty sure that if we looked at the data we could find similar patterns, where there are life changing events where there is disease or other things, or divorce," Dr Vogels says. "We don't do that, but we could actually recommend you books on divorce before you would even know that you would be interested."
Source: CMO Australia
Sign up for Computerworld eNewsletters.