Preparing to visit friends in the Berkshires for a ski trip, I went to Google Maps and began typing in the address to get a sense of where their weekend cabin was. I had entered just two pieces of information, a house number and the beginning of a common street name, when I noticed that high on the list of search results being dynamically delivered to me was exactly the one I was looking for.
It pinpointed my friends’ full address in a tiny rural Massachusetts village. From two generic bits of data, Google had figured out exactly what I wanted to know.
Impressed? Boy, was I. But truth be told, I was also a bit unnerved. It was one of those moments that cause people to look up from their laptops and swivel their head around to see if they’re being watched.
I told this tale to friends and there emerged several theories on how Google nailed it. These included the obvious (“You must have entered the address into that computer in the past even if you don’t remember doing it”); the clever (“Maybe you checked the site of the resort near their place to look into ski conditions”); and the conspiratorial (“Google is reading your Gmail and Google+ friends list and knows what you’re up to”).
Regardless of the answer, the experience neatly sums up my view of data-driven target marketing. When done right, it’s both impressive and unnerving, although the latter probably won’t be as true for younger generations who grow up expecting their devices to be smart about who they are.
While privacy concerns are understandable, what most people care about at the end of the day is the price-value equation. They’re willing to give up personal information about themselves if it helps a marketer target them more effectively and efficiently. If you can use past behaviors and stated preferences to deliver a better experience and waste less time, go for it.
There can also be a kind of joy in feeling that a brand you’re loyal to seems to know who you are. Visiting a clothing site and having it recommend that perfect pair of boots can be as satisfying as walking into a corner café and having the waitress remember that you like your eggs over easy.
Following the price-value theory, brands that misuse or abuse the data entrusted to them by their customers risk wasting their time, turning them off or losing their trust. Done wrong, target marketing comes across as robotic and awkward — a reminder that an algorithm rather than a person is calling the shots.
“Scott, we have recommendations for you,” I can remember Amazon‘s site boasting before displaying a series of baby books based on my purchase of one for a friend who was a new parent. Similarly, if I hit the “Genius” button on iTunes while listening to a Johnny Cash song, why does the algorithm assume I only want to hear country music? It’d be more impressive if the database was able to surmise that Cash fans also love, say, Bruce Springsteen and Mumford & Sons and return that playlist.
Of course there are ways for consumers to make smart databases smarter by letting Amazon know that they really loved that Steve Jobs biography or that the baby book they purchased was just a gift and shouldn’t be counted among their personal reading preferences — but this requires a time investment that only the most devoted and loyal customers are willing to make.
Gee, all this talk of complex algorithms is a bit overwhelming. I’m heading out for the simplicity of the slopes. I sure hope the ski shop’s database will remember what level of skier I am and what size helmet and boots I wore last year so they can set me up instantly with the right gear.
