Where’s the line between useful personalization and creepiness?
Unfortunately, it’s not a simple answer. But as we amass more information about our customers and try to build closer relationships with them, we need to think like a creep.
Chad White of ExactTarget just published a nice post that explains 6 different degrees of email personalization. We’ve come a long way from simple mailmerge fields that insert a first name! As we collect customer data, we can increasingly personalize communications based on what products they have (or have looked at), where they are, what similar customers have purchased, and so on.
As a customer, though, you may have gotten the occasional targeted email that just feels too close to home. You know the one. An email for a product you’ve gone back and looked at three times, but never actually added to your cart. Or a chiding email from a health insurer suggesting wellness tips for “women your age” that sound suspiciously like your mom’s advice.
When the personalization is dead wrong, it can be humorous, as when a 20-something friend somehow got profiled as a senior citizen in need of dentures, hearing aids, and death planning services. For months, he received several such mailed solicitations each week, all obviously targeted at someone 40+ years older. When they abruptly stopped, he assumed he’d finally been marked dead in a database somewhere.
But when the personalization is close enough that it’s plausible, or eerily accurate, many people pause. They may begin to rethink their relationship with the company, and how much data they share. I recently upgraded my Android phone to a newer model with Google Now services. It’s figured out where I go on certain evenings each week and now offers traffic updates, and it knows I follow the Blackhawks, so it offers me scores and updates. It scans my email and shows tracking information for packages en route. While much of this is useful, I’ve considered disabling some of these services, as they’re getting a bit too intimate.
As marketers, we need to keep this in mind as we personalize communications. Offering tailored messages based on past purchases or interests is great, but don’t be too perfect in the personalization. Tell customers why they’re getting such offers, such as Amazon’s, “Other customers who purchased X have also purchased Y.” But as Target learned when it sent a teenage girl coupons for prenatal products – before she’d told her family she was pregnant – we should tread the “creep” line carefully, especially when dealing with predictive data. Target now adds in “decoy” coupons that don’t fit the profile (such as wine glasses next to infant clothes) that make the offers seem more random and less perfectly targeted.
Have you seen examples of personalization that gives you the creeps? Share them below.