Web Masters Episode #36: Jeff Tarr

You might think using computers to find dates is a new thing, but that’s actually not true. Way back in 1965, an entrepreneur named Jeff Tarr used an IBM mainframe to build a computer dating service and, in the process, laid the foundation for the online dating industry. Get the full story on this episode of Web Masters:


Operation Match - Online Personals Watch: News on the Online Dating  Industry and Business

Aaron Dinin:

I want to share the story of a Harvard undergraduate who launched a revolutionary idea from his dorm room, that involved using computers to help people connect with each other in new and more personal ways. The idea was a hit. Within months it had rapidly spread to other college campuses around the country, spawning a lucrative business in the process and turning the founder into a celebrity, seemingly overnight. Now, I know what you’re thinking. You already know this story. It’s the story of Facebook and Mark Zuckerberg. But actually that’s not the story I’m talking about. The story I’m talking about began in 1965.

Jeff Tarr:

In those days there weren’t computers. You had IBM machines, IBM was the computer. And we had a card sorter when we started this situation, a friend of mine and myself only to meet women, we had no aspirations of getting filthy rich. And we started this then on Saturday night, when we, in theory, both were alone. I wrote the questionnaire and it had 75 questions. Most of the stuff was not very worthwhile, but we thought that wasn’t professional enough. We said, I’ll have the participant answer once for themselves and once for their ideal date. And we had 150 questions out of that, and my partner was very anal, so he laid it out like one of those very professional. I had been on a TV show called Password in New York and I was lucky enough to win $450. And that was my mad money. And I put in my 450 and he put in his money and we started. And we thought it would be a lot of fun. And we thought it would take us at most 10 hours a week. Well, it took us each 10 hours a day.

Aaron Dinin:

That was Jeff Tarr, founder of a company called Compatibility Research Inc, which as I already mentioned, he launched while a student at Harvard in 1965. Compatibility Research Inc, ran a service called Operation Match. And it was for a couple of years, the world’s most popular computer dating service. That’s right. Well, before match.com and OkCupid and Tinder, there was Operation Match. Are you ready to hear the story? Let’s get dialed in.

[INTRO]

Aaron Dinin:

Welcome to Web Masters. I’m your host, Aaron Dinin and I teach entrepreneurship at Duke University. This podcast explores the history of web based businesses by speaking with some of the Internet’s most impactful early innovators. We’re going to look at something a little different on this episode, because as you already heard, our guest wasn’t building his business online. However, I promise it’s all related, and I’m going to tell you how. But before I do, I want to take a moment to tell you about our sponsor.

Web Masters exists in part, thanks to the generous support of our partner Latona’s. Latona’s is a boutique mergers and acquisitions broker that specializes in helping people buy and sell cashflow positive internet businesses and digital assets. That includes things like Shopify stores, Amazon FBAs, SAS apps, domain portfolios, and yes, even online dating websites. If you have a revenue generating internet business, and you’re interested in selling it, reach out to the team at Latona’s because they have tons of experience helping people, just like you find the perfect buyer. Alternately, if you’re interested in being one of those buyers, then be sure to visit latonas.com where you’ll find listings for all the internet businesses they currently have for sale. That website again is latonas.com, L-A-T-O-N-A-S.com.

Lots of the stories we explore here on Web Masters focus on the immediate predecessors to contemporary internet companies. For example, we’ve explored the story of Friendster, the social network that proceeded Facebook. And we also talked about AltaVista, the search engine that preceded Google. This tendency to focus on the predecessors of contemporary internet companies shouldn’t be too surprising if you stopped to think about it for a second, after all the entrepreneurial process is really just a continuous cycle of leveraging technologies to solve old problems in new, better, and more efficient ways. In few places is that cycle more obvious than dating or perhaps the better word is mating because if we zoom out, not just on human history, but all biological history, we see that procreation is fundamental to the success or failure of a species. As a result, humans are always looking for ways to make mating easier and online dating happens to be the most recent iteration.

But we didn’t just jump from pre-internet ways of dating to swiping left and right on Tinder, there’s been a steady evolution of the matchmaking process in the digital age. So since in our previous episode, episode number 35, we talked with Sam Yagan, founder of OkCupid, I wanted to explore what came before that. Turns out the history of digital dating goes back a lot further than I thought. In fact, as early as the 1960s, even before the internet existed, people saw the potential of computers to help solve the challenges of finding soulmates.

Jeff Tarr:

At the time, let’s say you’re in college and you’re dating, girls, especially want to have a ring, their senior year, a ring by spring. The dating happened and it worked, but it didn’t work. In those days, they had mixers and a bunch of guys, a bunch of gals went to a place and they said, hello and they asked them to dance and you get to know them. But the mixers really didn’t work that well because you needed to know a little more about them and they need to know a little bit more about you. And so in theory, what would work is you could have a little background on these people, find out what they majored in, what their interests are, what they hope to accomplish in life. There was a need for some more information about the people you’re dating and vice versa.

Aaron Dinin:

Notice how Jeff is focused on what people need. Specifically, he says there was a need for more information about the people you were dating. Identifying needs is one of the first steps in a successful entrepreneurial process. It’s even of defined stage, the need finding stage and formal entrepreneurial discovery processes like design thinking and bio-design. Simply put, if the thing you’re creating doesn’t address a real need in the market, people won’t have any reason to use it and it’ll fail. As you heard that wasn’t the case for Jeff. Jeff saw a need. And while he didn’t have a way of connecting people via a website and helping them learn more about each other through an online dating profile, it’s not like he didn’t have any technologies at his disposal. For starters, he had paper and pencil, meaning people could still record and then share information about themselves.

So that’s exactly how Jeff started. He and his co-founder created a dating profile questionnaire containing 150 questions that he distributed to college students around the country. People would fill out the questionnaires and mail them back to Operation Match where the founders had discovered that computers, even at that early stage in their evolution, were a great way of managing and comparing large data sets.

Jeff Tarr:

It’s very obvious to us with computers, even back then that we can memorize a hundred things, but a computer could memorize a million things and you never can tell what would the match would be. And we asked all these kinds of questions and we did use some of the questions. You couldn’t do it all in your head. You really had to have it written down. And the computer was really, even though at that times, it was unusual, as the best way to keep the records, just enormous records. We wanted more information about the people and it had to be done with a computer. If you had written it all down, you would have probably gone through all the trees in America. The computer was the way to do it.

Aaron Dinin:

But computers were expensive room-sized machines back then, right? I mean, it’s not like you had one sitting on your desk in your dorm room, so how did you have access to the computers you needed to do the matching?

Jeff Tarr:

What happened was IBM, again, in those days, the computer was called an IBM machine. They had a division SPC, I think was initials. And they rented time on their IBM’s. So you had to have the information key punched into a card, and then they charge you so much per hour for using their machines. And machines were used quite a bit in those days, say 9:00 to 5:00 Mondays through Fridays. But if you use the machine Sunday mornings, Saturday nights between 1:00 and 3:00, it cost you very little. So we ran those machines between 1:00 and 3:00 and it worked. The information had to be key punched in. So we also hired people to key punch it. But once we had the cards, we had to sort them on the IBM machine. And we rented the time at Sunday mornings. It was very a hand to mouth kind of operation, initially.

Aaron Dinin:

And were you a computer guy back then? How did you program the matching process?

Jeff Tarr:

No. Very good question. Let me go into that from the computer technology point of view. I took some courses in Harvard that were supposed to be the tough courses, and I didn’t do very well, but one guy in there, he knew how to computer, how to make a program. So I paid this guy a hundred dollars to make out the program. And I also told him, I give him 10 dates. So I paid him both in cash and con and it worked out pretty well. But then a month after we got the results off, I got an eight page single space letter from a guy at the Harvard math department, going for his PhD. His name was Kirby Baker, an 8 page, single space, telling us how we should do it and the way it should go. So I called him up immediately and hired him.

Our software worked out real well. And I know now he became a professor of mathematics at UCLA, I think. And I think he retired recently, but it was very sophisticated and it was much beyond my capability, but I was interested in these kinds of things and it worked.

Aaron Dinin:

Okay, so you weren’t doing the coding yourself, but are you able to give us a sense of how the matching algorithm worked? What kinds of data points was it using to match people?

Jeff Tarr:

It wasn’t sophisticated like now, because if the people had to be in the same area, men wanted women their same age or younger, the women would want the converse, that many people they were whatever religion, they wanted someone for the same religion. So when you have say 2000 people in an area and you cut it in half and half and half the pool is like a hundred at most. And from then on, we had all kinds of questions, but it really didn’t matter. We just mashed them up.

Aaron Dinin:

So people just trusted the matches from the computer and that’s why it worked?

Jeff Tarr:

There was enough I think excitement of a computer that when people used computers, you thought that computers were right. Little did you know that garbage in, garbage out.

Aaron Dinin:

Yeah. We’ve definitely learned that lesson about computers since then. So once you ran everyone’s data through the computers and got their matches, how did you connect people with each other?

Jeff Tarr:

We got the matches and we guaranteed they would get at least six matches, but some people who had the right answers got up to a hundred matches. On the matches, so we’d give their names and their phone numbers. And in those days, most of the calls were made by the male, but not all. And it would go do their mailing address, it wasn’t email in those days.

Aaron Dinin:

Even though, as Jeff points out, they weren’t using email back in those days, hopefully you’re starting to see how the underlying principle of online dating is already there. Jeff’s customers would fill out questions about themselves and what they were looking for in a potential mate, the computer algorithm would sort through all that data to create a general pool of potential matches. And then the users would browse their potential matches to see if they liked any of them enough to go on a date. By the way, if you listened to our episode on OkCupid, you know that’s basically how their matching process works too. Really the biggest difference between today’s online dating and Jeff’s Operation Match processes the speed at which the communication happens well, that, and a lack of pictures.

Jeff Tarr:

We didn’t have pictures in those days. In fact, we had a question that we asked both for themselves and for their ideal date, how attractive are you? One, very attractive to seven, unattractive. Most of the people, of course, answered very attractive. Well, you and I know that not everyone is attractive to everyone.

Aaron Dinin:

Well, people still do the same things on dating websites by posting ten-year-old pictures or perfectly angled selfies and other stuff like that. It’s not like online dating is somehow perfect because it has pictures, right?

Jeff Tarr:

When you started getting pictures involved, you can obviously get a good picture, but it wouldn’t be the same thing as just saying you’re very attractive or moderately attractive. And we had a number of questions like that. And it wasn’t really a good thing. In fact, as you’ve got me thinking about it, one of the first mistakes we made was the first question was sex and not to be cute, the answers were male and female, but a number of women answered the first question, one, male, they just made a mistake. They just said male. So it turned out that one lady from Vassar who answered the question, the first question, wrong, she got matched out with her roommate, just turned out and it was a mistake that we had made asking that question first. So later on, we asked that question, what sex are you? And we asked that question like six or seven, but any business, anything you do, you make changes, and hopefully it improves.

Aaron Dinin:

All the things Jeff has been talking about so far identifying needs, solving them with new technologies, making mistakes and iterating quickly to fix them. Those are all things critical to businesses today. I mean, he might as well be talking about an internet startup launched six months ago, rather than a computer dating service from the 1960s. Which goes back to my earlier point, that entrepreneurial innovation, isn’t a discrete event, it’s a continuous iterative cycle. We can see that cycle playing out in the dating space. The basic functions of dating have been largely the same since well, presumably back in caveman days, but technologies have helped compensate for their inefficiencies, which mainly were related to who you could meet. In caveman days, you only knew the people in your immediate vicinity. By the time Jeff is running Operation Match, you’ve got students busing around to different campuses for mixers and parties in order to expand their candidate pool of potential mates.

Jeff and his team further expanded that pool of people someone could meet beyond their immediate vicinity, thanks to their computer dating service and internet dating services would eventually come along to help people cast an even wider net. Of course, as any user of online dating knows, in order for the services to be valuable there has to be a large number of people using them. The same was true for computer dating, which meant Jeff, like all great entrepreneurs, had to constantly be hustling for customers. However, unlike online dating sites, which notoriously have a heavy imbalanced toward male users, that wasn’t the case for Operation Match, which actually skewed slightly toward female users.

Jeff Tarr:

It turned out that the first time we got, I think it was 52% women. And I know when we got them, we went out to, one of my friends and I drove out to the Western Massachusetts and all these girls schools like Mount Holyoke. And I thought we were like Lewis and Clark going out. And we met these girls and we would talk about what we’re doing is a big social experiment. And a lot of girls out from those schools applied. One, there weren’t many boys schools around and two, they were interested in doing it. And so they would all get together and all of them would do it. So say from one dormitory in Mount Holyoke, there would be a hundred people on the dormitory and probably 80 would have done it. So we got a lot of ladies from that section.

And we ended up one summer, the second summer we were in the business, a hotel called Grow Singers. It was located in New York. They wanted to use the service and we charged them for it, because they have normally, they had singles weekends, but they had more women applying for singles weekend than men. So by having this computer dating for a singles weekend, they got a lot more women because they wanted to try it. And two, they ended up with a higher percentage of men and it worked for them at least as far as getting traffic, I don’t know what their results were in terms of the matches. And we told people, we were finding dates, not finding mates, but quite often, especially the women, we were finding mates for them.

Aaron Dinin:

So it sounds like most of your early customers came from the Northeast around where you were located at Harvard. How did people find out about it?

Jeff Tarr:

We got all kinds of publicity. First got the publicity because I called up all the local schools papers. And I told them what we were doing is a social experiment. And I told them if they got any people to do it, it costs $3. And that was big money in those days because $3, we’d give them 10% of their revenue. So sure enough, within about two or three weeks after the questionnaires were out, we got 7,800 responses. And so we were growing in money.

Aaron Dinin:

Keep in mind that 7,800 responses at $3 each would have been $23,400 in 1966, which is roughly equivalent to $200,000 in today’s money.

Jeff Tarr:

In the first year we ended up, we had the publicity and the following spring, we ended up with over a hundred thousand applicants.

Aaron Dinin:

Assuming those applicants were all paying $3 each, that’s $300,000 in 1966, which would be $2.4 million today. Not bad for the first year of a business started in a dorm room. I think even Mark Zuckerberg would be impressed.

Jeff Tarr:

And some of them, oddly enough, got married and some are still married. And I’ve told them all, it’s the law of large numbers. And in fact, in those days, if a computer said that was your ideal, you believed it. So anyway, that business, I was active in it for two and a half years. I went to no classes my senior year, I was an honors student so my tutor worked for me. And then I spent half of my senior year in New York and moved to New York with someone who had worked for, with me. And it was a big business for us. And we were lucky enough that in 1966 Look Magazine, which is not around anymore, on the same thing, did a Valentine’s issue about computer dating.

Aaron Dinin:

FYI, around that time Look Magazine would have had around 7 million subscribers.

Jeff Tarr:

We were on the cover, not myself, but this gorgeous woman whose name was Shelley Hack. She later became a famous model and a very nice looking guy as their ideal date. And the picture was taken in front of the Yale computer system, but it was really an IBM 1401. And so from February after that magazine came out for about a year after, we did wonderful things, but it was a fad and it didn’t work near term, nothing like Match or OkCupid countless others. But it was a great experience for me. When I started, I was 20, I couldn’t even sign the legal documents, I had to have my parents sign that. It’s a great experience and I think what works for almost everyone, if you find something you’re interested in, something that you have a passion about, it’s good to get involved with it, learn how to attack different kinds of problems. And attack isn’t really the right word, it’s not really an attack, it’s how you’ll be very comfortable with different kinds of situations. I think the big thing is just to have the experience.

Aaron Dinin:

As you heard Operation Match, wasn’t a long-term venture, it lasted about two years. But to be fair in those two years, Jeff became what we might think of as one of the earliest tech entrepreneurs, celebrities appearing on popular TV shows at the time like Johnny Carson, To Tell the Truth and I’ve Got A Secret. And even back then celebrity entrepreneurship had its pitfalls.

Jeff Tarr:

One thing that wasn’t fun but did happen and I’ve looked at it as a way to handle business. When I started the business, I had a phone number and my name was listed. Well, sure enough, after we started computer dating and a number of dates, weren’t very successful, the mothers of the people who applied would call me up and give me hell. So I stopped having a listed number for a long period of time.

Aaron Dinin:

But despite having to deal with angry mothers, Jeff’s notoriety and experiences with early computer algorithms had some valuable benefits later in his career. Specifically, after a brief stint in Washington, Jeff made his way up to New York City where he combined his high school actuarial experience with the lessons he learned about computers during his time with Operation Match. And he applied them to arbitrage trading. In fact, he became one of the pioneers of algorithmic trading on Wall Street.

Jeff Tarr:

I was lucky because I had some math ability and I had had this experience in Operation Match. So I set up a computer, which for then was kind of radical. And I looked at all these deals and Bourges, tenders and all those kinds of deals. And I worked out a little formula where they were in the process. If they had a definitive agreement, if they needed any trusts and how long it took. And we put in a program and so we had a hundred different deals. And once a day, my program calculated the deals, the spread, the absolute spread and what the expected value was. And it worked pretty well. And once a day was enough. Now, I think these days it’s instantaneous, but it worked out very well for me. And for the first seven years in the business, I doubled my income each year.

Aaron Dinin:

And Jeff eventually spun that work into another business based off of the algorithmic trading that he ran until he retired in 1995.

Jeff Tarr:

We had a partnership called risk arbitrators associates, risk arbitrators partners, and that shows you how creative we were. And for 14 years we ran it. We averaged 30% a year. And for the first period of time, until Reagan came along, I also featured doing tax arbitrage. So during that process, I would create at the end of the year, say making 30% on average, but we would show ordinary losses. If you were in the highest bracket, there was 70% and long-term gains the following year. And long-term gains were taxed at 30%. So you would make 40% if you just broke even, it was just unreal, but it worked out very well for me.

So after 14 years, I started that business in 1981 and in 1995, 25 years ago, I decided to quit because I had enough money and one kid was already in college and the other was about to go. And I talked to my wife and I said, “I’m not interested in being the richest guy in the graveyard, I want to have more adventures.” So for the past 25 years, I’ve been having adventures. Some have worked, some haven’t worked. Talking to you is one of the adventures.

Aaron Dinin:

Well, based on everything Jeff shared about his life, I think it’s safe to say being interviewed for a podcast about internet entrepreneurship wasn’t one of his most exciting adventures, but I know I really enjoyed hearing his story and I hope you did too. For me, it was a great reminder that the technologies and companies around us aren’t as revolutionary as we might sometimes like to think. And just as importantly, they’re not the end of the entrepreneurial process, they’re just another stop along the journey. I’d like to thank Jeff Tarr for taking the time to speak with us and share the story of Operation Match. If you enjoyed the episode, please take a moment to write a review on your favorite podcasting app and share this episode with a few friends.

Also, don’t forget to subscribe so we can let you know as soon as the next episode is released. If you have questions, comments, or thoughts about the episode, share them with us over on Twitter, we’re @WebMastersPod. I’m on Twitter too @AaronDinin. That’s A-A-R-O-N D-I-N-I-N. I also write lots of articles about startups, entrepreneurship, and business on medium.com. You can find those by searching for my name over there. A quick thank you to our sponsor, Latona’s for helping make this podcast possible. Remember if you’re interested in buying or selling an internet business, be sure to visit latonas.com. And a thank you to our sound engineer, Ryan Higgs, for his help, with editing. And finally, thanks to all of you for listening. Our next episode will be coming soon, but until then, well, it’s time for me to sign off.