Eharmony 19.95 for 3 months

9-709-424 J U L Y 1 , 2 0 0 8

________________________________________________________________________________________________________________ Professors Mikołaj Jan Piskorski and Hanna Hałaburda and Research Associate Troy Smith prepared this case. HBS cases are developed solely as the basis for class discussion. Cases are not intended to serve as endorsements, sources of primary data, or illustrations of effective or ineffective management. Copyright © 2008 President and Fellows of Harvard College. To order copies or request permission to reproduce materials, call 1-800-545-7685, write Harvard Business School Publishing, Boston, MA 02163, or go to www.hbsp.harvard.edu/educators. This publication may not be digitized, photocopied, or otherwise reproduced, posted, or transmitted, without the permission of Harvard Business School.

M I K O Ł A J J A N P I S K O R S K I

H A N N A H A Ł A B U R D A

T R O Y S M I T H

eHarmony

Greg Waldorf, the CEO of eHarmony, was in his car driving down the Interstate 10 Freeway after a day-long meeting with eHarmony’s senior leadership team. The sole purpose of this October 2007 meeting was to decide how the company should address recent competitor actions. After many deliberations, Waldorf’s executive team was able to identify four strategic options. Now, Waldorf and Greg Steiner, the President and COO, who was sitting next to him, were debating which option the company should pursue.

As the two whizzed down the car pool lane, passing cars stuck in traffic, they reflected on eHarmony’s success. This online personals site targeted marriage-minded individuals and offered a unique product which combined an extensive relationship questionnaire, a patented matching system and a guided communication system. Despite charging a premium for its services, eHarmony had experienced phenomenal membership growth while its competitors stalled. As a consequence, it was able to increase its paying membership base to slightly less than a half of its largest competitor, even though it entered the market six years after they did.

The success of eHarmony did not go unnoticed. From the beginning, competitors had been copying some of the company’s product features and closing the price gap. More recently, Match, eHarmony’s biggest competitor, had increased its advertising expenditures by 80 percent. Some of the increase was aimed at reviving Match’s sagging growth. However, most of it was spent on supporting the growth of Match’s new dating site, called Chemistry, which like eHarmony was a match-making service. It utilized different matching criteria and methodology, and was priced roughly 10% below eHarmony. To make matters worse, free personals sites and online social networks were exploding in popularity, challenging the business model of paid online personals.

As the two approached downtown Los Angeles, the carpool lane came to an abrupt end, leading them into a dense traffic jam. They could not help but wonder—was this also eHarmony’s future, or would one of these four options provide them with a new on-ramp to fuel growth and profitability?

Marriage Markets

Although the institution of marriage had differed substantially through time and across cultures, at the beginning of the 21st century, it remained one of the most central social institutions throughout the world. Created by a contract between two people, usually one man and one woman, marriage conferred special state, social, or religious privileges. It allowed a couple to share wealth; it regulated inheritance arrangements; and it legitimized sexual relationships and reproduction. It also regulated

For the exclusive use of N. Saaty

This document is authorized for use only by Nabih Saaty in Spring 2014 BUS 612A-03 taught by Kamal from February 2014 to August 2014.

709-424 eHarmony

2

political and commercial ties among families from which the couple descended. Spouses were expected to remain exclusive to each other and, traditionally, couples followed a strict division of labor, with women responsible for childbearing and men engaged in outside labor. Although this arrangement often led to significant power inequalities in the relationship, couples were expected to stay married until one partner died, with divorce being rare. Some cultures even regulated patterns of re-marriage, expecting, for example, a widow to marry the brother of her late husband.

In some cultures, couples were betrothed to each other by their families even before they became teenagers. Other cultures allowed the process to take place later, but still with significant assistance from family, close friends, elders, or even astrologers. In some instances, third-party matchmakers were employed to find advantageous mates. Even as modernity progressed, people were expected to choose their spouses through an elaborate process, often overseen by parents. In the U.S., the concept of a potential couple meeting privately at specified times and places to get to know each other started among middle-class teenagers in the 1920s.1 However, such encounters remained relatively formal through the 1950s. Cultural changes brought more freedom in the 1960s, with the first TV show focused on dating appearing in 1965.2 By this time, most individuals found spouses through friends, colleagues, or family members, or at work, church, or school.3 By the early 1980s, the concept of marriage had evolved from a functional partnership to an institution based on “love, sexual passion, or even close friendship.”4 Indeed, Americans no longer believed that the purpose of marriage was to have children. Now, 70% wanted “their spouse to make them happy.”5 This shift in expectations made potential spouses less of “search goods,” with clearly identifiable characteristics, and more of “experience goods,” which could only be “judged by the feelings they evoked rather than the functions they performed.”6

Modern Marriage Markets

These cultural changes, coupled with several economic factors, had a substantial effect on the marriage market in the 21st century (see Exhibit 1). First, only about 16% of U.S. singles (about 7% of the adult population) reported currently looking for a romantic partner. Men were less active than women in going on dates, but men and women were equally likely to report that it was difficult to meet people.7

Second, the marriage rate had reached its lowest point in recorded history. In 1890, men married at a median age of 26, while women typically married at 22. By 2004 the age had increased to 27 for men and 26 for women (see Exhibit 2). The later marriage age was partly attributable to cohabitation before marriage. Only 25% of those who cohabited got married within five years of first moving in together, but almost 60% of those who married in the early 2000s had lived with their future spouse for two years before getting married.8 Between March 1995 and March 2005, cohabitation rates had risen from 2.9% of adults to 4.7%.9 Although the cause was uncertain, couples who lived together before marriage were 50% more likely to get divorced than non-cohabiting couples.10

Third, divorce rates had changed substantially. Starting from a low rate in the 1950s, the divorce rate rose rapidly in the mid-1960s, reached its peak in 1981, and had been declining since.11 The divorce rate was lower for people who were well-educated, wealthy, and religious, and for those who married someone of the same race after the age of 25, came from an intact family, and did not have a child before marriage (see Exhibit 3). A general increase in the age at first marriage and the educational attainment of married couples were often used as explanations for the overall decline in the divorce rate.12 Despite the trend, the American marriage market experienced more “churning” than most developed nations (see Exhibits 4). Whether this was caused by financial disagreements, infidelity, communication problems, sexual problems, mid-life crises, addictions, career strains, and/or emotional or physical abuse, divorce put an end to about 45% of marriages, with most

For the exclusive use of N. Saaty

This document is authorized for use only by Nabih Saaty in Spring 2014 BUS 612A-03 taught by Kamal from February 2014 to August 2014.

eHarmony 709-424

3

divorces occurring in the fourth year of marriage.13 Americans were also more likely to re-marry, with a median time between divorce and second marriage of 3.5 years, and with men more likely to get re-married. 12% of men and 13% of women had married twice, but only 3% of either group married three or more times (see Exhibit 5).14

Personals Industry

The oldest surviving record of a personal ad was published in the Collection for the Improvement of Husbandry Trade in 1695. It read: “A Gentleman about 30 years of age, that says he has a Very Good Estate, would willingly Match Himself to some young Gentlewoman that has a fortune of 3,000, or thereabouts.” As time went on, personal ads became more common, particularly where men faced a shortage of women. An ad for “any gal that got a bed, calico dress, coffee pot and skillet, knows how to cut out britches and can make a hunting shirt, knows how to take care of children can have my services till death do us part” would not be unusual on the American frontier.15 Women, however, were often not allowed the same social rights. Helen Morrison, for example, a lonely English spinster, placed a personal ad in 1727 in the Manchester Weekly Journal; the ad caused so much stir in the community that she was committed to an insane asylum for four weeks.16

As dating became more popular in the twentieth century, individuals began to advertise for potential partners more frequently. Initially people took out personal ads in newspapers and magazines. These ads usually gave a short blurb about themselves and the type of partner they sought. Increasingly sophisticated communication and recording technology led to the creation of telephone personals. The advent of video tapes in the 1980s brought images to personal ads, which could be cataloged by dating services and exchanged among spouse seekers. These services were complemented by professional matchmakers, who charged up to 100 times more for their services, but who used individual interviews and background checks to personally select dates for clients.

Internet personals

The advent of the Internet expanded partner seeking opportunities. The first online personals firms, started in 1992, met with lukewarm response. Internet savvy people were the most likely to try the new services, but these individuals preferred to meet future spouses through AOL chat rooms or free bulletin boards. By doing so they could avoid the social stigma of being on an online personals site, which at the time was considered appropriate only for the truly desperate. The use of the Internet for quick hookups made the stigma even worse. Even as paid online personals sites became more acceptable, the level of customer dissatisfaction remained high. Frustration and disillusionment affected the serious relationship segment the most. Successful customers got married and left the site; few would publicly admit to having met online. Those who could not find a match remained in the market, vociferously complaining about the problems with online dating. As a consequence, the online personals market was growing very slowly, reaching only $40 million in 2001.17

By 2005 the situation had changed dramatically, with 16 million people claiming to have visited an online personals site at least once.18 Growth was attributed to changes in attitudes toward online dating; social acceptability grew as more people knew someone who had met their spouse online. This change in attitude was so extensive that it even benefitted offline matchmakers, whose number increased by 25%, despite charging $1,500 to $10,000 for their services.19 Initially, online daters were younger than the general population, but quickly the 40– and 50–year-olds became the fastest- growing segment of the market (see Exhibit 6). Significant growth also came from people who signed up at the request of their families. This category could comprise as much as 20% of paying members, although this number varied depending on the target market of the site. Of those who had ever used personals websites, 45% had never been married; 31% were divorced, separated, or widowed; and

For the exclusive use of N. Saaty

This document is authorized for use only by Nabih Saaty in Spring 2014 BUS 612A-03 taught by Kamal from February 2014 to August 2014.

709-424 eHarmony

4

23% were married.20 Seven million people had gone on dates with people they had met through an online personals site, with 40% entering some type of a long-term relationship. Of the 2.2 million marriages that took place every year in the U.S., reportedly 120,000 occurred between people who had met on an online personals site.21 Industry insiders claimed a much higher rate, believing that nearly one-fifth of marriages were initiated through an online encounter.

After 2005 the percentage of Internet users who visited online personals sites fell from 20% to 10%, but the decline in visitors did not translate into a decline in paying subscribers, allowing the industry to grow to $900 million in 2007. Many believed there was still substantial room for growth in the market (see Exhibit 7). Only 37% of people who declared themselves as looking for a relationship and with access to the Internet had actually been to a personals website. However, once someone subscribed to a service they tended to be repeat customers of online personals sites. Consequently, observers believed that the industry could double in size by 2012.22

eHarmony

eHarmony was founded by Dr. Neil Clark Warren and his son-in-law, Greg Forgatch, in 1998. The company officially launched in August 2000, six months after receiving $3 million from a Houston investment firm, Fayez Sarofim & Company.23 After earning his Ph.D. from the University of Chicago, Dr. Warren spent 35 years as a practicing psychologist focused on marriage and family relationships. During that time, Dr. Warren had also written nine books on love, marriage, and emotional health and made numerous appearances as a speaker at conferences and seminars as a relationship expert. Warren had also been dean of the graduate school at Fuller Theological Seminary. With decades of professional counseling experience, Dr. Warren saw the opportunity to help people with mate selection. He focused eHarmony on singles seeking a serious relationship—a segment of the market that had not been well served until then. Unlike other websites, eHarmony also introduced matching on the basis of long-term compatibility.

In the early years of eHarmony, Dr. Warren worked very hard to spread the word on eHarmony’s unique offering. He appeared on a wide range of TV and radio programs including local news programs, two appearances on Politically Incorrect with Bill Maher, and Good Morning America. A major breakthrough occurred in August 2001, after the company was featured on a Christian radio program affiliated with Focus on the Family, an evangelical organization, which had published some of Warren’s books. eHarmony’s focus on serious relationships resonated well with faith communities—a market segment also untapped by traditional dating sites—resulting in over 100,000 registrations in just a few weeks. By early 2002 registrations had grown to over 300,000, allowing the firm to break even that year and become cash-flow positive the next.24

Subsequently, the company expanded its TV and radio advertising campaigns to appeal to a broader audience.25 As a result, by early spring 2004, eHarmony hit three million registrations, at least 40% of which were considered active users. The company was famous for being able to convert these active members to paying members three times more effectively than the industry average.26 Later that year, Technology Crossover Ventures and Sequoia Capital invested in the company. In 2005, Greg Forgatch stepped down as CEO, while continuing to serve on the board of directors. Greg Steiner, previously the Vice President of Operations, stepped into the President and COO role. The Board led a search for the CEO position, with Greg Waldorf, the founding investor and Board member, taking the role temporarily and leading the search process. Ultimately, the Board asked Greg Waldorf to remain on as CEO permanently.27 Under Steiner and Waldorf’s leadership the company grew to 230 employees, approximately a half of whom were in customer service.

For the exclusive use of N. Saaty

This document is authorized for use only by Nabih Saaty in Spring 2014 BUS 612A-03 taught by Kamal from February 2014 to August 2014.

eHarmony 709-424

5

Product

eHarmony distinguished itself from other personals sites by offering a tightly integrated system that encompassed a Personality Profile, which fed into a matching algorithm, which then led to a Guided Communication system.

The Personality Profile initially contained 436 questions, which the company pared down to 258 (see Exhibit 8). The questions covered 29 basic measures of compatibility, including personal lifestyle preferences, communication style, values, beliefs, attitudes, personality traits, family background, birth order, energy level, intelligence, spirituality, special interests, and future aspirations. eHarmony developed the Personality Profile internally. “We went through the whole process of establishing content validity and developing a voluminous set of items to assess the domain. We then had different people look at the items and pare them down and give them to focus groups, and then to larger groups to get some initial estimates of reliability. We were quite rigorous in every step of developing the questionnaires of personality, values, and interests, which were the three primary constructs” recalled Dr. Galen Buckwalter, Chief Scientist at eHarmony. Having developed the instrument, Warren and his team surveyed over 2,000 couples before the website was launched.

The Personality Profile formed the central part of the Relationship Questionnaire, which was required of anyone who wanted to become an eHarmony customer. The questionnaire was free, but the original version with over 450 questions required at least 1.5 to 2 hours to complete.28 More recently, the questionnaire has been abbreviated to about 250 questions, which takes about 45 minutes to complete. Men were less likely to complete the questionnaire once they started. Waldorf explained: “Since it is hard to sign-up, the eHarmony person self-selects. There is a shared sense of investment to be part of eHarmony. It says, “I’m serious.” Despite the length of the questionnaire, more than 14 million people completed it in the first seven years of eHarmony’s existence.

The completion of the Relationship Questionnaire did not automatically guarantee access to the service. The company declined to sell memberships to as many as 20% of its potential customers. The majority of the time, the company declined to sell memberships because people were already married, followed by those who were underage, or had been divorced more than three times. This policy was challenged in a lawsuit, Claassen v. eHarmony, filed by a married man under California’s Unruh Civil Rights Act accusing eHarmony of discriminating against him because he was not offered a membership due to his marital status. When the Plaintiff registered for the site, he was married and living with his wife , which was inconsistent with the company’s policy that all of its members be “free of relationship commitments.”

Since its inception, the company has declined to sell memberships to at least one million people who sought to become paying customers, costing the company an estimated $10 million per year. Waldorf believed this was a sound business decision: “We leave a lot of short-term value on the table, but the idea is to keep the quality of our pool really healthy.”29 The final distribution of members reflected the geographic distribution of the U.S. quite well, with a slight skew toward less populous areas. Approximately 60% of eHarmony users were women, who generated slightly more than two- thirds of visits to the website.30 In 2007, people 45 years and older constituted the fastest-growing segment of users for the company.

The company also did not offer its singles matching service to women seeking women or men seeking men. This choice has attracted considerable positive and negative attention. Given eHarmony’s limited resources and the relatively small size of the same-sex market, the Company did not believe it made sense to enter that marketplace. eHarmony did not exclude the possibility that it would engage in the research needed to establish statistically valid and reliable matching models for

For the exclusive use of N. Saaty

This document is authorized for use only by Nabih Saaty in Spring 2014 BUS 612A-03 taught by Kamal from February 2014 to August 2014.

709-424 eHarmony

6

the same-sex market in the future. The company also faced a second lawsuit, which it believed was baseless, also under California’s Unruh Act, Carlson v. eHarmony, filed by a lesbian woman accusing it of discrimination because it does not offer services for people who want same-sex matches. Other companies that specialize in same-sex matching that do not offer opposite-sex matching have not been targets of similar litigation.

Matching Algorithm After an applicant had completed the Personality Profile, the answers were fed into the matching algorithm. If the applicant had the potential to be successfully matched, then she or he would be able to purchase a subscription. The matching algorithm had been developed by a team that included Warren and Buckwalter, among others. Although most psychological literature on the topic suggested that “opposites attract,” Buckwalter recalled that the team was “convinced that successful relationships were almost universally characterized by a high degree of similarity, particularly in areas like intellectual ability and emotional stability. We considered similarity to be more important for personality characteristics, then values, then interests. We also thought that agreeableness and emotional stability were very important.” Having identified the algorithm, the team set out to conduct a set of studies of married couples to test and refine it. Buckwalter believed that “the ideal design would have been a longitudinal study, but we could not do that within the constraints of the business model. So the decision was made early on to study married couples, and to make the assumption that if we got really good at predicting satisfied and happy marriages, that we could apply that to singles.” Between 2000 and 2004, eHarmony did numerous rounds of matching algorithm validation with over 4,000 couples. The results were very encouraging: the algorithm could predict to a high degree of accuracy whether couples would end up in the top quartile of the Dyadic Adjustment Scale, a tool used by researchers to measure long- term relationship happiness.31 The company secured a patent for the algorithm in May 2004. Although critics believed the patent amounted to little more than a marketing gimmick, further company research revealed that successful couples who had met on eHarmony were significantly happier than couples who had met “in the wild” or through other online means.32

When eHarmony made a match between two people, both sides were informed and could review basic demographic characteristics about the other, along with answers to some of the “about me” questions (cf Exhibit 8). Each party could then decide whether to start communicating. If one of them decided not to pursue the match, both sides were informed and the match was considered “closed.” Although the matching system took full advantage of the initial Personality Profile, it did not use information from the matches a customer decided not to pursue. “You don’t like that this person has a pet, or you’re a vegetarian but you keep being matched with hunters” reflected one eHarmony team member. Improving the ability of members to set screening preferences was a next generation feature in the works.

In order to communicate with one’s matches, a member had to buy a subscription, for which eHarmony charged almost twice as much as other online personals sites (see Exhibits 9 and 10). “Subscription is driven by how users like their matches. It’s almost like everyone has an implicit dollar amount they are willing to pay for a match. We can predict re-subscription rates based on the number of hypothetical pairings—how many matches are available for a person—so people who are easier to match are more likely to re-subscribe,” commented Waldorf. He continued: “People see the value of a good match when they see an attractive profile, of which an important piece is the photo. Philosophically, we are trying to help people get to know each other more deeply and at a more fundamental level. But we had to face facts that members who have photos are between 9 and 15 times more likely to receive a message. Now, 80% of paying subscribers have a photo posted.”

The distinction between paying and non-paying members created some problems for the company. Non-paying members were still matched to others, meaning that a paying member could

For the exclusive use of N. Saaty

This document is authorized for use only by Nabih Saaty in Spring 2014 BUS 612A-03 taught by Kamal from February 2014 to August 2014.

eHarmony 709-424

7

initiate an interaction, leading to the conversion of non-paying members to subscribers. However, some paying subscribers felt frustrated by this system. “If I am a paying user and you are not, I can send you a message and you might never respond because you don’t want to pay. And then I don’t know—did you not like me? Did you even see me? I want some kind of feedback. Otherwise, it feels like I am investing all this time to start communication but after all this investment, I never hear back,” reflected an eHarmony user. To address this concern, eHarmony was considering adding a feature to allow users to see which messages had been read.

Guided Communication Unlike the matching algorithm which was the result of intense study, Warren came up with the initial design for Guided Communication, which led potential couples through a set of questions before allowing them to communicate directly. Based on his years as a clinical psychologist, Warren believed that people, when left on their own, would “gravitate to the most superficial questions, like sports or activities, not to what really matters in relationships.” Waldorf echoed the sentiment: “you need to give people a way to talk about deep issues such as children and ideal locations. Guided Communication allows for such information-seeking without the social stigma.”

Guided Communication comprised three distinct activities. First, each member in the pair was asked to choose five easy-to-answer questions from a list provided by eHarmony, and send them to their match. Questions such as, “If you were taken by your date to a party where you knew no one, how would you respond?” would be followed by multiple-choice answers, such as “(a) Stay close to my date, letting him/her introduce me, (b) Find a quiet spot and relax alone, (c) Strike out on my own and make friends, or (d) Ask my date if I could skip the event.” Once both parties answered, they moved to the next stage where they were asked to exchange their personal list of “must haves” and “can't stands.” In the final stage they were asked to exchange three open-ended questions to allow for more detailed description of both parties’ values. eHarmony provided some sample questions, such as “What person in your life has been most inspirational, and why?” or “Tell me about your closest friend. How long have you known them, and what do you like best about them?” But members could also write in their own questions. Once this exchange was completed, and a message from Warren was displayed, the two parties could move into Open Communication. During Open Communication the pair could send emails to each other, exchange photos, and prepare for their first meeting. A potential couple could then decide when, where, and how to meet in the offline world if they wanted to pursue a relationship. Buckwalter commented, “When people meet in person, they have all this collective history and discussion, and it’s almost like they know each other right off the bat.” The company estimated that on average a successful subscriber took four to six months to get matched to someone they would eventually marry.

At any point in the process, either party could “close” the match and cease any further communication. Given the number of opportunities to drop out, only 20% to 30% of matches ended up in open communication. Upon mutual agreement, the pair could also elect to Fast Track their interaction directly to Open Communication, without going through Guided Communication. Only 10% of eHarmony members used Fast Track. Though men had most frequently requested the Fast Track feature before its introduction, they were less likely to use it, because within the standards of eHarmony it could be seen as intrusive. The company recently allowed users to state their preferred communication method, leading to a tripling of Fast Track requests. Members who used Fast Track communicated with more potential matches and renewed their subscriptions more often, leading to higher lifetime value. “These people come to us because they want to meet somebody, and they want to interact to determine whether they would like to get to know their matches better. The first step is emailing back and forth, then talking on the phone, or meeting in person. Our members who are able to do this more quickly have a higher satisfaction rate and stay on the service a lot longer” commented a senior Product team member.

For the exclusive use of N. Saaty

This document is authorized for use only by Nabih Saaty in Spring 2014 BUS 612A-03 taught by Kamal from February 2014 to August 2014.

709-424 eHarmony

8

The results of this integrated system were astounding. eHarmony commissioned a study, conducted by Harris Interactive between August 2004 and August 2005. The study found that on average, 90 eHarmony members married every day in the United States as a result of being matched on eHarmony.33 A follow-up study in 2007 found that on average, 236 eHarmony members marry every day in the United States as a result of being matched on eHarmony, representing a stunning 2% of marriages in America.34

Marketing

To support growth in its membership base, eHarmony invested substantial resources into marketing. The company aired its first radio commercials in 2002, with TV spots following in June 2003. Soon thereafter eHarmony stumbled on a very successful campaign, which paid for itself in increased subscriptions within a week! It featured testimonials from happy couples who had found love through eHarmony, in short upbeat segments. A senior marketing team member believed that the commercials clearly communicated that “eHarmony is about serious relationships and helping people find lasting love. Nothing could convey this more authentically than the couples. A lot of our early growth came from bringing people into the category who previously wouldn’t have considered online dating, because it seemed too sleazy, too casual.” Although there were at least 100 different executions of the campaign, the couple was always placed against a white background and accompanied by Natalie Cole’s “This Will Be” song. The spots often mentioned the Personality Profile, the 29 dimensions of compatibility, and automatic matching. Warren, a “wise, experienced doctor,” was also featured prominently in the ads, although it was never explained that he was a psychologist.

In 2007, the company was considering adjusting its four year old signature campaign to appeal to those interested who have so far shied away from eHarmony’s service. The new campaign featured spots focused on one couple, shot in their real environment in a documentary style, highlighting that couples love story. Any changes to eHarmony’s successful marketing formula had to be introduced with care; the company had earlier experimented with telling the story of lonely individuals who were afraid that they would never meet anyone, followed by a happy conclusion. These commercials backfired, because “the initial ten seconds spent recounting the ‘problem’ just flared up anxiety in people and dampened our response rates… so we went back to focusing on the end benefit,” explained a team member.

The successful advertising formula made eHarmony one of the few online companies that made offline marketing work and pay for itself. But with marketing expenses reaching as much as $80 million per year, firm profitability depended on efficient customer acquisition. eHarmony was extremely judicious in its use of advertising dollars, avoiding any pure brand-building activities and focusing on direct-response marketing. The marketing team worked only with advertising and media buying agencies, “who really understood direct-response advertising… This means they are incredibly diligent about buying media at lower rates – both by locking in good ‘upfront’ deals and searching for last-minute remnant inventory.” As a result, eHarmony advertised only on national cable networks and avoided broadcast television, where prices where higher because local stations allowed for greater targetability than national cable. Match, eHarmony’s close competitor, used broadcast TV and aired its commercials during primetime shows, paying $20 to $30 per thousand impressions—at least four times more than cable TV prices.

With three-quarters of the marketing budget spent on TV and radio, eHarmony spent the remainder on online marketing, including paid Internet search and banner ads. The latter had high visibility and got decent click-through rates, but the conversion rates were low, which made banner search more expensive than paid search in terms of customer acquisition.

For the exclusive use of N. Saaty

This document is authorized for use only by Nabih Saaty in Spring 2014 BUS 612A-03 taught by Kamal from February 2014 to August 2014.

eHarmony 709-424

9

Research & Development

Since its inception the company had made significant investments in R&D, culminating in the opening of eHarmony Labs in 2007. Staffed with five research scientists, the Labs were tasked with studying the biological, sociological, and neurological underpinnings of love. The Labs boasted over 2,000 square feet of clinical space, including several rooms set up with chairs and couches. In this setting, couples were observed as they interacted to examine relationship dynamics.

eHarmony was also pursuing research on physical attraction. Buckwalter noted that “physical attraction plays a large role in the initial meeting but is a very poor predictor of long-term success. After the initial meeting, people find themselves attracted to others with whom they share common values. But if we can give people matches with whom they’re also going to feel this ‘click’ factor, they’re more likely get to the point of actually developing a relationship.”

Finally, the company has invested substantial resources into a five year study of 400 couples. Couples were enrolled during their engagement and followed through their marriage and subsequent life stage transitions, such as pregnancy and childbirth. Early results already suggested that “the biggest adjustment of every marriage is the birth of the first child.” Now, the study sought to identify what characteristics and behaviors in couples predicted successful transitions. The company believed that its team of research psychologists was uniquely positioned to identify specific insights and turn them into products and services that could become new businesses for eHarmony.

Competitors

Competitor Types

Paid Do-It-Yourself sites were the most common type of online personals site. Some of eHarmony’s direct competitors, such as Yahoo! Personals and Match, were in this category. In contrast to eHarmony, these sites put up very few barriers to joining. Individuals were required to provide basic information about themselves, such as their age, location, gender identification, and sexual orientation. Optionally, they could provide a short blurb about themselves and provide a set of pictures. For most heterosexual sites, men were more likely to sign up than women. As soon as the short registration process was complete, individuals could specify the criteria for their partner search. The website instantaneously provided them with a set of profiles and pictures of individuals who matched their search criteria. Members could then browse profiles to identify their own choices for a match. Some sites allowed unlimited browsing, and all required a subscription to communicate with other members. On average, 5% of those who signed up became paying subscribers, with men more likely than women to sign up. Paying members had little loyalty and, on average, belonged to at least three other personals sites, either as visitors or paying members.35 Sites tried to develop some differentiation by offering compatibility tests, or by conducting background screenings to weed out sex offenders, felons, and married people, but most of these features were ineffective or easy to replicate. Perhaps the only meaningful distinction was the advertising spend.36 One firm, True, spent $90 million on advertising in a single year—an amount that exceeded its revenues numerous times, without much long-term impact.

The process of looking for a match was fairly time consuming. By some estimates, online daters spent nearly seven times as long searching for potential partners as interacting with them. Heterosexual sites reported fairly skewed patterns of interactions, with a select group of women being inundated by messages from men. Patterns of interactions between the two genders could be easily predicted by basic characteristics. A study of speed-dating found, for example, that “men avoided women whom they perceived to be smarter than themselves. When women were the ones

For the exclusive use of N. Saaty

This document is authorized for use only by Nabih Saaty in Spring 2014 BUS 612A-03 taught by Kamal from February 2014 to August 2014.

709-424 eHarmony

10

choosing, the more intelligence and ambition the men had, the better.…Women also exhibit a preference for men who grew up in affluent neighborhoods.…Women also put greater weight on the intelligence and the race of the partner, while men responded more to physical attractiveness.”37

Online interaction was designed to lead to offline meetings. However, researchers found that online daters typically “ended up going out with fewer than 1% of the people whose profiles they studied, and….those dates often ended up being huge letdowns.”38 Acknowledging the difficulties of dating in general, researchers found that “most participants were so dissatisfied with both online and offline dating that they would have preferred to stay home and watch a movie.”39

Free Do-It-Yourself sites had been rapidly growing in membership base. The growth defied the common industry wisdom which claimed that requiring people to pay to join a dating site served as an important barrier separating lurkers and casual daters from serious ones. Being free, these sites could attract members quickly, leaving it to the members to interact with the vast number of others to identify their own match. In addition to SinglesNet, the fastest growing among these was Plenty of Fish, founded by a 29-year old entrepreneur in Canada, Marcus Frind. Despite a terrible user interface, Plenty of Fish had become the most frequently visited dating site in the U.K. and Canada and had reached No. 4 in the U.S. by the end of 2007. Although Plenty of Fish only had half as many visitors as eHarmony, the website received 20% more visits than eHarmony did. Such user engagement on the site resulted in 900 million page views per month, allowing this one-person company to become one of Google’s AdSense top earners, with revenues close to $500,000 per month. A significant portion of this revenue came from referring people who were dissatisfied with the quality of matches on Plenty of Fish to more selective paid sites, such as Match or eHarmony. Not all free personals sites were equally successful, however. OKCupid, started by the founders of eDonkey (a free file-sharing service), was only beginning to catch up, despite having a more appealing interface and much better functionality. Some sites, such as Craigslist, did not even care about making profit, even though they possessed a significant market share of personals, particularly those targeted at casual hookups. Few doubted that the existence of these sites put a cap on how much the paid sites could charge for their services (see Exhibit 11).

Niche sites were also in competition with eHarmony. Waldorf referred to the niche sites as “community of interest sites,” which he said were “characterized by low customer acquisition costs. For the generalists, it is basically a customer acquisition story.” The gay and lesbian market was the largest niche market, attracting about a tenth of the internet traffic of the heterosexual personals sites.40 . Some sites catered exclusively to this market, while others added it on top of heterosexual matching. Sites targeted at African Americans, such as BlackPeopleMeet or Black Singles, were a close second, followed by faith-based personals site, which included sites such as JDate.com, ChristianSingles.com, Hindu-Dating.com, Muslima.com, and LDSsingles.com. Of these, JDate, targeted at a Jewish audience, was considered the most successful, with almost 10% of Jewish singles enrolled on the site. Other sites covered just about every conceivable demographic—some targeted at graduates of elite colleges, such as TheSquare, others focused on people with health conditions, such as HIV, Crohn’s disease, cancer, or diabetes. Internet users could also sign up on RichorBeautiful, HotEnough, or HotorNot, as well as SingleParentLoveLife, SweetOnGeeks, FarmersOnly, SugarDaddyForMe, VeggieFishing, and GothScene, not to mention MarriedButPlaying. Overall there were at least 850 different personals sites on the Internet. About the same number entered the industry every year, with most failing quickly without making a penny of profit.41

Online social networks provided an important substitution threat to the online personals industry. Large online social networks, such as MySpace or Facebook, did not have to incur the vast advertising expenses required to attract customers, instead relying on a “viral” process through which friends encouraged their friends to join. Such networks also held a significant advantage over

For the exclusive use of N. Saaty

This document is authorized for use only by Nabih Saaty in Spring 2014 BUS 612A-03 taught by Kamal from February 2014 to August 2014.

eHarmony 709-424

11

personals websites in the veracity of information they provided about potential matches. The online dating industry was plagued by people misrepresenting personal information—ranging from subtracting a year or two from their real age or increasing their height by a couple of inches, through posting 10-year-old pictures or losing 50 lbs. overnight, to outright fraud, by failing to disclose that they were married. Because people were connected to their friends in online social networks, however, they would find it hard to post a 10-year-old picture or forget to mention a spouse, without eliciting comments from friends. Online social networks were also attractive to people who were currently in long-term relationships and could not easily join a dating site without jeopardizing their current relationship. Under the guise of interacting with their friends on an online social network, they could scan the market for new partners. Indeed, some industry experts estimated that the most traffic was generated by men in relationships looking at women they did not know. Whether such actions translated into offline liaisons was still unknown. Finally, online social networks held a significant advantage over dating sites for gay men and women. Due to inherent search difficulties in this market segment, and the importance of compatibility between one’s partner and one’s friends, online social networks provided a valuable search tool.

Despite these apparent advantages, online social networks presented a number of limitations compared with online personals sites. First, online social networks did not make it easy to send a clear signal about whether one was interested in forming a relationship. Even though online social networks allowed people to indicate whether they had joined only to interact with friends or to establish a new romantic relationship, at least 40% did not indicate their marital status at all, making it hard to figure out whom to approach. Second, with more people concerned about the privacy of their information, an increasing number of profiles were becoming private, with access available only to one’s direct friends. This shift severely curtailed the usefulness of online social networks for meeting or finding out detailed information about people one did not know. Finally, online social networks were used less often than personals sites by people in their 40s and 50s, making them less useful for that segment of the population. Weighing the advantages and disadvantages, experts at Jupiter Research saw “no signs that the eruption of social networks has burned the paid online personals market.”42 Waldorf shared this view: “I don’t believe that social networking will hit us in ways that everyone expects….I think it will have less to do with pricing and more to do with user expectations and the user experience. We have a new cohort of users from MySpace and Facebook coming every day. Will they find the guided process to be a breath of fresh air or too constraining?” Others within the industry concurred, believing that social networks played in a different space and that among all the people meeting online, only about 5% had met through a social network site.