SXSW 2011: How to Personalize Without Being Creepy
Presented By
Hugo Liu
Chief Scientist at Hunch
@dochugo
Jennifer King
PhD Candidate at University of California at Berkeley
@kingjen
Mat Harris
CEO at BizGreet Inc
@matharris
Noah Weiss
Project Manager at Foursquare
@noahweiss
Vijay Ravindran
Chief Digital Officer at The Washington Post Company
@vijayravindran
Why is personalization important?
Life is too short to see random stuff. We have the ability and the data to show you relivant data, so why can’t we? But it is important that users understand where the personalization comes from and how you know this data about the person. If someone wants to know why, there needs to be a pathway to tell the user why they are seeing this. There also needs to be a way for the user to turn it off if they are uncomfortable with it. Users don’t read privacy statements, and they don’t read it when you change it. We need to take back privacy from the lawyers. Great relationships come from intimacy and customers want to be intimate with a brand. But it must be done right. This is as important as supply chain optimization in the 90s. It is affecting our economy.
What is the line between great and creepy personalization?
Trust is fleeting once you have violated it. Expectations is crucial. The user needs to have control over what is shared and what is personalized. And the privacy should be the most strict when the user starts. Start conservative to gain trust. “Don’t try to get laid on the first night.” But don’t hide your intentions. Personalization should be a dialog with the user.
Should personalization always be opt in?
Users don’t always understand what is going on behind the scenes. With a single facet model, the user might understand (Twitter, Foursquare) but when a user is being tracked across multiple sites, it is much more confusing and creepy, so these should be opt in. Product design is like designing a building. You wouldn’t make a building that isn’t easy to get out of, but there are a lot of online communities that are hard to opt out of. Opt in can be an opportunity to get the user to provide more data, or better customize their experience, if they understand the benefit they will see.
Customization vs. personalization
Customization is allowing the user what they want to see and what they don’t. Personalization is about using data to derive information about the customer.