In one brief scene from the 2002 movie Minority Report, Tom Cruise walks through the mall, where talking billboards scan his retinas, identify him, then call out his name to get his attention.

Check out this scene here:

This disturbing future vision has become closer to reality than we could have anticipated. Dozens of online companies have built data profiles on people that contain hundreds, or even thousands of data points.

The optimistic promise of personalization is a huge win for everyone involved – less waste from the company’s perspective, and more relevant messaging and experiences for the end user. However, when personalized experiences are not constructed with sensitivity, they can backfire because they creep users out.

Target’s efforts to acquire more customers by advertising to women they believed were pregnant had the unintentional outcome of inadvertently including at least one unwed teen.

When a personalized experience backfires, companies lose real brand equity. So, how do we create personalized experiences that will be accepted as helpful and beneficial by users? How do we avoid being creepy?

Gain Consent Through Opt-In (Don’t Force An Opt-Out)

The best personalized experience is one that users agree to, with a complete explanation of what data is being collected and why. For instance, Intel’s Museum of Me uses data from the user’s Facebook account to make a unique, museum-style video showcasing the user’s friends, status updates, photos, and other social aspects.

This creates an emotionally resonant message, and connects the customer with the brand deeply. It’s particularly challenging for a brand like Intel, whose chips are never seen by the end consumer, to create an emotional response. But, personalizing the experience creates an immediate connection for the customer.

Ad retargeting, on the other hand, faces backlash because products mysteriously follow users all over the web. The user doesn’t know why, so it is perceived as creepy spying.

Target solved this problem by sprinkling in deliberately mis-targeted ads alongside the targeted ones. For the pregnant women, this meant inserting ads for lawnmowers next to ads for unscented lotion and calcium supplements. But, this level of subterfuge only makes consumers more uncomfortable when it is (inevitably) exposed.

Mark Pincus and Mat Harris facilitated a dialog about this at SXSW, where the “black box effect” (unknown algorithms determining personalization behind the scenes) was seen as problematic and disconcerting.  In the session they moderated, audience agreed that consent was the optimal way to avoid being creepy.”

Define the Give-to-Get

Personalized experiences should offer real, tangible benefits to the user that are easily understood. What the user gives up, in terms of personal data, must be worth less to them than what they get in return.

For some users, that may be as simple as a $5.00 off coupon. But, for the highly valued customer, it will likely be something of much greater value, and sometimes, not something of monetary value. Think broadly about what you can offer a user, including status and reputation, which can’t be bought for any price. Dr. Pepper, for example, allows users to write their own declaration, and have it featured on the Dr. Pepper home page, in exchange for logging into the site and sharing some personal information. Dr. Pepper offers their fans a chance to show off their reputation by being featured on the home page – an honor that money can’t buy.

Align Personalized Elements to the Channel

With increasingly integrated customer management systems, the customer data that you have access to is the same whether the user is in store, on the web, or in a mobile app. But the data you should use to personalize the experience must align with the channel.

Yelp does a good job differentiating its level of personalization, depending on the device. On the web, your Facebook friends’ reviews of a business are prioritized, and you can browse the activity of your social graph. This makes sense because users are in evaluation mode – trying to decide which happy hour to attend.

But, on the mobile device, the experience is personalized primarily by current location. The user is out and about in the world, and needs an answer that is geo-targeted. By focusing on the needs of the channel, Yelp personalizes appropriately.

Create Experiences for Your Target Audience

Different people trust your brand differently. And, digital natives feel differently about privacy and data than digital immigrants. So, defining the audience that you’re trying to reach means that your experience will be resonant for the target market.

You should expect that someone, somewhere will be creeped out by your personalization experience. This is inevitable – we each feel differently about our need for privacy, and who we trust with personal data. But, if we align with the expectations of the target market, we will create value for our customers.

These potential complications may make you feel like creating personalized experiences is not worth the effort. But, it’s actually a necessity. With so many personalized experiences on the web, from Amazon’s recommended products; to Google+ impacting Google search results; to Twitter’s personal Trending Topics; the “Internet of Me” creates the expectation that the web will be customized. By not participating, you risk becoming irrelevant.

What challenges have you uncovered in creating personalized experiences?