"It's not the consumers' job to know what they want," Steve Jobs famously said. Jobs was talking about products that hadn't been envisioned yet—there was no way for consumers to know they wanted a phone that gave them directions until such a thing even existed.
But there's a broader sense in which consumers may not quite know what they want, says Shayan Zadeh, cofounder and co-CEO of Zoosk, which bills itself as the "smart way to date." As Zadeh has learned in running a dating site, there's a gulf between what people say they want, and what their behavior reveals about what they actually want. Oftentimes, the key to running a successful business may be watching what your customers do, rather than what they say. (At any rate, it's working for Zoosk, which recently revealed that it made $40 million in revenues in the first quarter of 2013.)
Zadeh offers what he calls an oversimplified example. There might be a user on his site who says he just doesn't go for brunettes—has no interest in them at all. But pay attention to how this gentleman acts—whose profiles he clicks on, who he chats up, and who he winds up dating or even marrying—and lo and behold, it turns out he'll make an exception here and there. And it was allowing for that exception that made his experience on the site. As with hair color, so with many other facets of a potential mate.
All this is quantifiable. Zoosk tracks a user's behavior, feeds this into an algorithm, and makes a more highly educated guess the next time it serves up potential matches. In essence, Zoosk has learned how not to listen to its users—how to know what they mean, rather than what they say. "This modeling we do is very similar to credit score modeling," says Zadeh. In the same way a system tries to answer the question, "What is the likelihood this person will pay her next bill?," Zoosk's modeling attempts to answer the question, "What is the likelihood this pair of people will hit it off?"
"We still do compatibility questions," says Zadeh, mentioning that dating site staple. "But those to me are just the icing on the cake. Who we actually show you is the most important factor, and we do a lot of work before we show you anybody."
Of course, applying Big Data to the Big Date is nothing new—various sites talk shop like this at the annual Digital Dating Conference, which Fast Company has covered before. But Zoosk's focus on deeds above words sets it apart—at least from most other dating sites. Zadeh points out that other businesses have long ago learned when to listen to the client, and when not to. He gives the example of real estate agents, who know that even when a client says he'll only take a south-facing three-bedroom in North Brooklyn, it's not unheard of for him to fall in love with that west-facing two-and-a-half-bedroom in South Brooklyn. "Agents mix and match," Zadeh says. "They'll give you 90% what you ask for, but they'll throw in another 10% that deviates from that."
Why don't people know what they want? I ask Zadeh. "It's not that they don't know what they want; it's what they perceive that they should want," he says. "It's a complex psychological thing. It's, ‘What do my friends expect me to go with?' ‘What does society expect me to like?' All these combined create a mask over what you truly like, and you might not even know you're interested in something until you see it."
Zadeh thinks that more businesses ought to use Zoosk's trick of watching what users do, rather than what they say. Though he cautions that you need to have an awful lot of data to do it well. Zoosk has 40 million singles in its system; between 10-15 million of them are active in a given month.
If you're reaching scale like that, it may be time to stop listening to your customers and the start watching them. "I'm sure Amazon is looking at behavioral information extensively, and if they're not, they should be," says Zadeh. Then again, he realizes there are certain obvious differences between the kinds of matches Amazon and Zoosk make. "Their problem is simpler than ours," he says. "A book can't turn you down."
[Image: Flickr user 4nitsirk]