Finding Love Online

Dating-Game Theory

By Bill Snyder

As first dates go, it rated four stars. The antipasto platter and pizza were good; the conversation was even better. Both spoke Italian, loved Mediterranean food, and had lived in Florence, so there was plenty to talk about. The chemistry? “I knew from her pictures that she was beautiful,” says Dan Beltramo, MBA ’94, “but when we met, it was love at first sight.”

His date, Laura Cheney, must have agreed; before long they were dating steadily. But the couple was a bit hesitant to tell friends they had met via eHarmony, a popular online marriage and dating site. Instead, they claimed to have bumped into each other at a local BART station.

Beltramo, who went on to found Vizu, an online market research and ad measurement firm, and Cheney, a physical therapist and graduate of the University of California, Davis, married on June 17, 2006, 18 months after their first date.

Not every online pairing has such a sweet ending, of course. Finding lasting love in the virtual world may be no less certain than in the real world. But millions of singles are turning to matchmaking sites in hopes of meeting Mr. or Ms. Right, and some entrepreneurs believe that both profits and better marriages will result. Whether it’s coincidence or some thing in the Palo Alto air, at least four such sites—eHarmony, Match.com, OkCupid.com, and SpeedDate.com—were founded, or now are managed by, graduates of the Business School.

WaldorfGreg Waldorf, MBA ’94, the CEO of eHarmony, was looking for a socially conscious business and joined the company because he shared its goal of promoting happy marriages and reducing divorce. Sam Yagan (at right, standing), MBA ’05, and Dan Abelon and Simon Tisminezky, both MBA ’07, felt online romance is one of the next big things and likely to be a commercial success.

Matchmaking sites have different approaches, ranging from psychological profiling and controlled studies used by eHarmony to the on-the-fly video meetups facilitated by SpeedDate. Some charge for their services, some are free at least temporarily, and some survive through advertising. What they have in common is a belief that the internet is a better place to find a mate, or at least a date, than the real world. “We don’t add the magic, people add the magic,” says Galen Buckwalter, who heads eHarmony’s research efforts.

Despite what you see in the movies, mate-finding has never been just magic. People have always turned to their social networks—tribal matchmakers, country clubs, churches, and work groups—to find a mate. The large category of social networks these days also includes internet sites such as Facebook and LinkedIn, which greatly expand an individual’s loose connections to friends of friends, or customers of customers. Just as the odds of finding an out-of-print book are greater now that websites like BookFinder.com are there to help, it would seem that the internet might make it easier for people to find better business partners, suppliers, contract manufacturers, and, yes, even romantic mates.

Indeed, Stanford sociologist Mark Granovetter in 1973 published one of the most quoted writings in sociology, titled “The Strength of Weak Ties,” which demonstrated the advantages of utilizing connections that exist beyond our day-to-day circle of friends and business associates. He found that professionals are more likely to get jobs through the weak ties of casual acquaintances than through the strong ties of close friends. The work has implications that go well beyond career building.

More recently, Michael Ostrovsky, an assistant professor of economics at the Graduate School of Business, has been studying the effect of the internet on various electronic commerce–related markets. While he hasn’t looked at matchmaking per se, he sees an interesting similarity. “In both cases, transaction costs are lower, and the market is suddenly thicker.”

A “thicker” market, he adds, means more choices. In electronic commerce, there are more buyers and more sellers; in matchmaking people, there are more potential mates, which means a better chance of a happy outcome.

In a virtual marriage mart, you might think that the typical shopper is a lonely, nerdy white kid, sitting in his dorm room or Silicon Valley apartment. Not even close.

Interviews with online dating execs and the relatively small amount of academic work focused on the topic suggest that users come from a broad demographic swath, with women generally outnumbering men by a small percentage, a wide spread of ages and religions, and significant representation of minority groups and the disabled.

At eHarmony, most users are aged 25 to 54, but the fastest growing segment recently has been people in their mid-50s and up, says Waldorf, an early investor who became the company’s second CEO. “It’s like America, people of all ages and races,” he says.

A 59-year-old woman showed up at one of SpeedDate’s first events and “wrote that her 85-year-old dad wanted to join,” said Dan Abelon, the site’s cofounder with classmate Simon Tisminezky.

How many people use these sites is hard to know, and how many find lasting love is even harder. Since eHarmony’s founding in 2000, some 18 million people have registered, though not all are active, paying users of the site.

When two Wharton researchers wrote about online dating in 2007, they cited data from Forrester Research suggesting that 16 percent of singles are using online personals. A September 2005 internet tracking poll by the Pew Internet & American Life Project and Princeton Survey Research Associates International suggested that 3 percent of internet users who are currently coupled first met online. And with 72 million U.S. households on the internet as of 2006—according to Gartner, a technology research and consulting organization—those numbers are obviously huge.

What’s the attraction? Users of these sites normally don’t talk about thicker markets and lower transaction costs, but they likely are drawn by the possibility of improving their odds or shortening the time involved in finding a good match.

In real-world market terms, people in college and graduate school find the shelves, as it were, crammed with merchandise—eligible singles. As they get older and enter environments that are not age segregated, the store shrinks, says Paula England, a professor of sociology at Stanford. “With the divorce rate so high, there are many people reentering the market without an easy way to meet people,” says England, who met her current love interest via Match.com, founded and then sold by Gary Kremen, MBA ’89.

That great first date enjoyed by Beltramo and Cheney wasn’t serendipity. The good match was—at least in part—the product of data crunching by eHarmony’s computers. After his classmate Greg Waldorf suggested he try the service, Beltramo spent a few hours filling out the company’s online questionnaire, which features about 250 questions focusing on 29 different personality attributes. The answers are the raw material plugged into a regression analysis that aims to provide couples with detailed profiles of only those the computers deem compatible.
“We’re looking for similarities; without exception they are the proverbial money in the bank when it comes to long-term relationships,” says eHarmony’s Buckwalter. But what about the idea that opposites attract?

“We say that opposites attract and then attack,” he jokes.

Buckwalter presides over a department that includes five PhDs, an elaborate observation lab in Pasadena, Calif., and a database of research built on more than 15,000 interviews and questionnaire data from the millions of people who have registered over the years and from married couples. He is a protégé of Neil Warren, a clinical psychologist and marriage counselor who studied more than 5,000 couples before cofounding eHarmony. At regular intervals eHarmony adds to its questionnaire database on married couples and then constructs a series of algorithms that attempt to correlate a potential couple’s likes, dislikes, and personality traits with behavior that predicts marital satisfaction.

The team employs regression analysis, a familiar tool in the social sciences in which a historical database is mined for correlations between individual causal factors and particular outcomes. Simply put, a regression analysis predicts the future based on past patterns.
If there is an 80 percent probability that a model can correctly put someone in the upper 25 percent of good relationships, eHarmony adds it to the system, Buckwalter says.

Ian Ayres, a former Stanford law professor now on the faculty at Yale, wrote about the use of statistical analysis by various matchmaking sites. He noted that True.com uses a regression analysis based on data from 99 relationship factors, while Perfectmatch.com uses a variation of the Myers-Briggs personality test, which classifies people into 16 basic personality types.

Perfectmatch and True.com look for complementary personalities, a type of matching that Buckwalter dismisses. “It’s hard to tell who is right,” Ayres wrote in his 2007 book Super Crunchers, “because the industry keeps its analysis and the data on which the analysis is based a tightly held secret.”

Stanford’s matchmaking alums have all been around the dot-com block. Sam Yagan is founder of a relatively new advertising-supported service called OkCupid.com, where users take personality quizzes and play trivia-type games. Each answer they give adds data points to their profile, making it easier for the system to suggest potential partners. Yagan recently launched an even newer website called Crazy Blind Date.com. As the name implies, it matches people with a minimum of screening and sends them off on real-world blind dates. The site is making matches in four cities and draws on OkCupid’s database of thousands of singles to pair people up with as little as a few hours’ notice.

Just 30, Yagan is on his third dot-com, and he’s spoiling for a fight with the big kahunas of the online dating world. “If I can build a site that is free to consumers, acquires customers at no cost, and is better than the subscription alternatives, I should become the biggest online dating site on the web,” he says. That’s pretty brash, but he previously built a solid online business by challenging CliffsNotes with SparkNotes, and then took on the record companies with eDonkey, one of the largest peer-to-peer file-sharing applications. And no, Yagan didn’t meet his wife online; he married his high school sweetheart, Jessica Droste Yagan, MBA ’05.

Waldorf, who has a penchant for socially conscious businesses, helped found Start Up, an East Palo Alto nonprofit that provided training for aspiring entrepreneurs in that struggling community. He and his mother, Toby Waldorf, founded Destination-U, which helps high school students find the right match in a college. Why move to couple matchmaking? “What we do has such an impact on people’s lives. The couples [matched by eHarmony] are very special to us,” he says. Waldorf is married, but the details of his family life, like his company’s profit margin, are private.
Not everyone is pleased by eHarmony’s mission. The company has been criticized and is currently being sued for not allowing people to look for partners of the same sex. “We don’t have a political agenda. We’re running a heterosexual matching site; because of the size, it’s the market we’re going after,” Waldorf says. He adds, however, “That could change in the future.”

Business, meanwhile, is strong. Sales this year are expected to hit a record $200 million, and the company is developing new lines of business and working to improve the accuracy of its matchmaking. The research staff is busy studying the elusive quality of chemistry between people, and eHarmony expects to add “the click factor,” as gut-level attraction or chemistry is called in the internet world, to its service sometime in 2008, Buckwalter says.

The way Dan Abelon sees it, people know who they like—in an instant. “It only takes a few minutes to know if there is chemistry,” says the SpeedDate cofounder, who met his current girlfriend “the traditional way—through a blind date.”

As he did market research while a student in an entrepreneurship class at the Business School, Abelon heard the same story again and again. “People spend a lot of time [on profile-driven dating sites], but when they meet they find there’s no chemistry.” And that, Abelon decided, meant there was an opening in the market for a different kind of player. SpeedDate invites participants to fill out a very basic questionnaire, and then sets them up with a series of quickie web video dates. At the end of three minutes, each partner votes thumbs up or down. Then it’s on to the next partner. (If both partners agree, SpeedDate supplies contact information for a real-world date.)

The site’s first test run attracted 230 people who went on a total of 1,100 three-minute dates in an evening. Within a few weeks, the site had arranged more than 20,000 dates across the country.

Will such speed-dating lead to long-term matches? It’s hard to know, but there is some evidence the format can affect the outcome.

It’s no news that when it comes to picking a mate, women tend to value traits like intelligence, sincerity, and earning potential, while men are looking for, well, babes. That’s obviously an oversimplification, but research generally supports it.

However, when Business School professor Itamar Simonson and colleagues from Columbia and Harvard looked at speed-dating, they found that under the time pressure of a speed date, women go for looks rather than the deeper traits they told the researchers they cared about before the test dates began. “Women do not say that appearance is particularly important to them, but it is, particularly in the context of speed-dating,” Simonson says.

Given that it’s much easier to judge someone’s looks in a four-minute encounter than to grasp their deeper qualities, the results make a lot of sense. The study also found that women tend to be choosier the more options they have. One possible explanation: Women may invest more emotional energy than men in each date and therefore not want to solicit dates from as many potential partners. What that says about the potential for success of sites that give women hundreds of choices in potential mates isn’t known but could be very interesting.

It’s too soon to draw broad lessons about the significance of online matchmaking, but the enormous number of people attracted to the sites makes one thing very clear: The stigma attached to the computer-assisted date is vanishing rapidly. After all, it was only a few years ago that Beltramo and Cheney didn’t want to tell their friends how they had met. And the early days of eHarmony were very rough; the fledgling company almost foundered in 2001.

But with the internet changing so much of our lives, why should the human heart be exempt?