There’s a real gap between what people think they want, what they tell us they want, and what they really want or really respond to,” says Grant Langston, CEO at dating website giant Eharmony. “And if you talk to enough people, you come to see that people don’t even know this.”
Luckily for them, there’s artificial intelligence.
AI and machine learning is the heart of their business, Langston says, and it’s super complicated.
“If you do a better job for people in this business, they leave you,” he explains. “That’s one of the challenges for our industry, right? But we’ve taken the approach that if you leave and you’re happy and you found the love of your life, you’re going to tell people, and it’s going to be the best marketing that we can do.”
The day you start with Eharmony, you’re just a number, placed in a bucket defined by the answers you’ve supplied in your onboarding questionnaire — 140 questions that determine what type you are, based on how others of that type have responded historically.
Once you begin to interact with your matches, from the folks you reach out to, to those you respond to and the ones you ignore, the algorithm starts to learn who you are, what you want, and what you don’t want (or who you don’t want). And the longer that you interact with the website, the better the algorithm understands you — better than you understand yourself.
“We tend to think of ourselves in certain ways, and we may even intellectually think we like certain things,” Langston says. “But when presented with a choice on a dating site, you’re not making it intellectually, but at a much deeper level of the brain.”
So as the algorithm continues to learn and give you more of what you didn’t even know you were interested in, you build brand affinity, continue to interact with the site and provide more data, and continue to improve your chances of success.
All matches are compatible, Langston says — they’re not interested in offering an incompatible candidate just to satisfy a need you think you want.
“We decided long ago we didn’t want to do that — and the better our technology responds to you, the better your experience,” he says. “We’re not going to find everyone the love of their life. It turns out that people don’t even expect that. We’ve done enough research to understand that people want to go on dates with people that seem like good candidates. And if we do that for them, then they leave satisfied.”
The company’s journey into the heart of artificial intelligence shows that AI isn’t something that necessarily has to be baked in from the start. Eharmony has been in business since 2000, and still maintains a data center with racks and racks of hard drives, Langston says, which has tremendous implications for how they build software and how they can improve their service while it’s running.
“You fix the plane while it’s in the air,” he says. “Our process of learning about you has gotten much better over all these years. That’s filtered in to changes in the way the sites work, from the top to the bottom.”
And they’re continuously looking at ways to improve the effectiveness of their system of matching affinities and connecting people who want to fall in love, he adds, but as ever in the artificial intelligence sphere, it’s always a balancing act.
“Tomorrow we may find we can do the most amazing thing, but we always have to come back and say, is this good for helping people fall in love with someone that’s good for them?” Langston says. “When the guy comes in and says, ‘Grant, guess what, we can do this now,’ I have to be the one who raises my hand and says, ‘I know that’s cool. I know that’s fun. But does it help people fall in love?’”
In the end, the best kind of AI is the most unobtrusive, he says.
“If we’re doing our job well, you don’t know anything that we’re doing — it just feels like romance to you,” he says. “But we’ve got a bunch of guys in white coats behind the curtain engineering this stuff and trying to make it happen. That’s the magic of our business.”
To learn more about how artificial intelligence can bring the magic for your own brand, how to create personalized engagement that keeps them coming back for more, and how AI can make you a tactical magician, don’t miss this VB Live event!
Source: VentureBeat
To Read Our Daily News Updates, Please Visit Inventiva Or Subscribe Our Newsletter & Push.
