When DreamWorks screened an early version of its 2001 film Shrek for test audiences, the female lead, Princess Fiona, made children scream in fear. It wasn't because the animated character turned out to be an ogre but because she looked too much like a real person. The studio’s animators, in a fight to keep up with Pixar, had created the most realistic human characters ever before seen -- in fact, they were too real for human comfort.
This phenomenon is known as the “uncanny valley,” the point at which a leap forward in technology outpaces a human’s ability to cope. Today, entrepreneurs who market their services online are confronting their own version of this. The idea of their data being analyzed, targeted, retargeted and marketed to based on contextual data can make consumers squeamish or even angry, unless marketers handle the situation with care. (According to the Harvard Business Review's website, contextual marketers might be considered those using "the power and reach of the internet to deliver tailored messages and information to customers at the point of need.")
As commercial real estate and other late adopter industries launch into their first-ever contextual marketing programs, smart marketers recognize that approaching data science like a kid in a candy store is not the most effective way to win over audiences. So here's a few lessons from other industry leaders’ hard-won wisdom to define three basic principles that will help businesses better make use of big data:
1. Be radically open about the data collected and how it's used.
At the core, what makes people squeamish about data science isn’t a fear of the technology itself, but a fear of what humans will do with it. On the corporate side, Facebook, which is not known for being an open platform, has developed such a reputation for privacy violations that the launch of its Messenger app was met with a huge wave of revulsion, even though the permissions being requested were really pretty standard. A University of Vienna study showed that half the users who have left Facebook have done so because of privacy concerns.
Radical openness about the data being collected is not just an ethical choice. It’s smart marketing. If companies aren’t proactive -- radically open -- about defining and communicating privacy priorities, they leave the door wide open for a reporter or misinformed user to shape the company’s privacy narrative. And when that narrative includes allegations of hidden secrets or deceit, it immediately destroys user confidence in the company.
2. It's possible to do more when users share the data themselves.
Imagine a woman showing up for a blind date and learning that the person she's going to meet has already ordered her favorite dish and asked the band to play her five favorite songs. Theoretically, the date is providing a perfectly targeted experience, but if she didn’t give the person any of this info, she's not going to react favorably.
When a company frames its messages around specific info that customers didn’t knowingly provide, it can be just as off-putting.
Contrast that with Foursquare, whose new app openly asks users to select their likes and interests (examples: artichoke pizza, great deals, mellow jazz), so the app can serve up location-based recommendations. Suddenly, users are experiencing great contextual messaging while still feeling in control of their data, and the uncanny valley is avoided.
Netflix has also used user-reported preferences to great success for years. Songza.com pushes this boundary even further. Songza asks users for highly personal details such as their mood in order to serve up playlist options.
Facebook, by contrast, this summer found itself facing an Federal Trade Commission complaint after a revelation that the company was altering users’ newsfeeds as a social experiment with emotions. Users were not involved with the data collection process. Plus they didn’t even know it was happening and this created a storm of anger toward Facebook.
3. Resistance to data sharing may soon erode.
In Japan, where realistic androids have been part of the culture for many years, the visual uncanny valley has diminished considerably. Robots are used for everything from modeling clothing to caring for the elderly, and consumers have generally come to accept them.
In the United States a similar erosion of resistance will take place as people begin to experience real benefits from contextual marketing and data science.
In other words, the more U.S. companies use their data to create valuable solutions for everyday life, the more people will accept it. Consumers today might experience a gut-level cringe at the idea of an artificial intelligence that can recognize and analyze their Instagram photos, but companies like Google and thought leaders like Jeff Bezos and Elon Musk are investing heavily in it, making huge bets that people will accept it someday. So it’s important that U.S. companies don’t stop researching or innovating based on the limits of society’s comfort levels at any given point. They just need to be extremely sensitive about what technologies used, how and when.
Looking at these three principles, it becomes clear that navigating data’s uncanny valley is about thinking like a scientist, a good citizen and a marketer at the same time -- finding the intersection between what’s possible, what’s ethical and what’s effective. Just as it isn’t smart to hit consumers with every new data technology available, it also isn’t smart to hide capabilities or refrain from dreaming them up in the first place.