This article originally appeared in Re/code.
Back in 2004, “The Polar Express” sounded like a sure-fire holiday hit: A big-screen version of a beloved children’s book, brought to life through cutting-edge live-action performance capture. But the movie flopped. During the post-mortem, many suggested that the film’s standout feature was actually one of its core problems: Audiences were put off by the overly lifelike animations. As one critic put it, “those human characters in the film come across as downright … well, creepy. So [it's] at best disconcerting, and at worst, a wee bit horrifying.”
Welcome to the “uncanny valley.” The term comes from robotics professor Masahiro Mori, who described how people react positively to increasingly humanlike representations, until the point at which they get too close to a real human being, when they suddenly become repulsive. It’s this phenomenon that makes zombies and clowns give us the heebie-jeebies.
Sadly, in this age of Big Data, the term is frequently used to describe data-driven marketing that feels invasive, overly familiar, or just plain wrong. By now you’ve probably heard about the dad who found out his daughter was pregnant when Target sent her coupons for baby items. Or the other dad who opened his mailbox to find a direct-mail piece from OfficeMax addressed to “Mike Seay, Daughter Killed in Car Crash, or Current Business.”
We each have our own uncanny valley. It could be a stack of junk mail that swells with each birth and death, marriage and change of address. It could be a row of display ads that calls out with the ghosts of emails, searches, and status updates past.
I found mine after I threw a baby shower for a friend and started to receive marketing materials from a local hospital’s maternity department. The uncanny valley looks different for everyone, but you know when you’ve taken that sudden step into the abyss.
Ultimately, data-driven marketing is trying to “fake” intimacy, to recreate true relationship through a series of data points. But because this approximation is so imprecise, because so many targeted ads miss the mark, the accurate ones only serve to highlight this farce. Indeed, our latest research at Communispace found that consumers experience targeted ads as an unpleasant mix of accurate, annoying and alarming. As tech writer Farhad Manjoo put it, “targeted Web ads are too dumb to be useful, and just smart enough to make you queasy.”
Of course, if you can do a good enough job of disguising what you know, you might avoid suspicion; you might even see some short-term gains. But you will be venturing only further into the uncanny valley, further away from your customers and not closer to them. When Big Data promises big results at the push of a real-time bidding button, where do we go from here?
The good thing about valleys is that they can be crossed. Relationship can be earned, and it starts with treating consumers with respect, as true individuals and not targets. When there is a fair and explicit exchange of data for value, consumers are quite willing to share; to be tracked, even. That’s why they like personalized deals from brands with which they do business, but hate targeted promotions from strangers. We found that 74 percent of consumers think it’s okay for companies to offer personalized coupons based on their purchase history, but only 13 percent think it’s acceptable for any company to buy or sell personal data in order to personalize.
A few tips to help brands avoid the uncanny valley:
- Use data to personalize the entire brand experience. Leverage customer data for more than just advertising. Allow customers to create personalized defaults to shape their ongoing interactions with you. Customize your website — content, UI, microsites — for different consumers. By allowing consumers to customize their own experience, they become partners in the personalization, rather than passive recipients.
- With Big Data comes big responsibility. Just because you have information, doesn’t mean you have to use it. Know your boundaries and use common sense. Don’t infer around sensitive topics (e.g., pregnancy, weight loss). With prospects, don’t become overly familiar, or you’re just a stalker. Marketers need to use their own empathy and judgment. Remember your human side when you’re at the wheel of the big-data machine.
- Help data work for customers, not against them. Think about new ways to use data — to serve customers, rather than yourself. Show them how their behavior compares to others, or provide tools for tracking, analyzing, and even exporting their transaction history. Be transparent and let consumers have some control over their data, and they might just share even more.
Above all, stop being a stranger. Stop looking through the window, and instead ring the doorbell. Better yet, unlock your own front door. Emerge on the other side of the uncanny valley and say hello to your consumers. You might just discover a new breed of creative, collaborative partners. For real.
Katrina Lerman is a senior researcher at Communispace. Reach her @communispace.
Please download our report, pass it on, and join the conversation below.
