Home » How I Gamed Online Data to Meet My Match: Amy Webb at TEDxMidAtlantic (Transcript)

How I Gamed Online Data to Meet My Match: Amy Webb at TEDxMidAtlantic (Transcript)

Amy Webb – American futurist

So, my name is Amy Webb and my story starts six years ago. I was in not the best relationship although it started really well and I thought that the person I was with was the person I would wind up marrying and our relationship came to a somewhat abrupt end. And I was devastated.

But I looked around at my friends and my family members and the people that I knew and the people that I admired and realized that there were a lot of people who were divorced; there were a lot of people in pretty bad relationships and a whole bunch of people who were generally not happy. And I thought, “What’s wrong with all of us?” Right? We’re all smart people. There’s got to be something wrong. Maybe there’s something wrong with me?

And so I consulted my friends and my family and my grandmother who — in between mahjong games — said, “Play the Field!” “Stop being so picky!” “You got to go out with everybody!” And most importantly, “True love will find you when you’re least expecting it.”

Now, we’ve all heard this before and you may not come from the same background that I do but I’m a numbers person. I’m a data and math person. And “least expecting” my way in the true love made no mathematical or scientific sense. But online dating did. I had a whole bunch of people suggesting me that I try online dating websites and to me that seems a lot easier — going trough that sort of data — than trying to find somebody in real life.

So I decided to create a couple of profiles. I went to Jdate, which is a site for Jewish people and I went to match.com. Now the problem was that I had a very full schedule and the last thing I wanted to do was to sit down and start answering a whole bunch of questions as if I was taking some kind of Cosmo quiz — this is gonna shock all of you but I am not a Cosmo quiz kind of woman.

So I did what any woman in my position would do. I copied and pasted from my resume, including bullet points! So I had worked tirelessly to make sure that my resume was spectacular and I was quite proud of all of my professional accomplishments so I listed all of those along with some other highlights like the fact that I spoke fluent Japanese and I was also fluent in CSS and Javascript.

Now, obviously we all know now that this was not a good idea but I want to take you back to 6 years ago. You see, work was really important to me and I feel very much that my professional life defines who I am also, and perhaps most importantly this bullet point resume that I copied and pasted onto my online dating profile didn’t prevent me from finding dates in fact, these algorithms, at Jdate, Match had stuck me with plenty of people. And we went on truly awful dates. For example there was Steve, the IT guy.

Now, online he seemed spectacular. He said he was 6 feet tall, muscular. He was a foodie who liked to cook and an IT manager, who loved gadgets. The problem, as I soon realized was that Steve in real life was a very short 6′ tall and quite stocky he did like gadgets and he was an IT manager but one of the things that I realized when we went out was that he liked to order lots and lots of food that was very very expensive. So he ordered all kinds of dishes the most expensive bottle of wine on the menu and when the bill came he actually pushed it towards me on the table.

Now, I am a modern woman and I am totally OK with paying the bill sometimes and splitting the bill sometimes. But I didn’t order all of that and the bill came up to what at the time, was an entire month’s rent for me so reluctantly I pulled up my credit card and I thought, “That’s it, you know, we’re done. I’m leaving.”

Outside the restaurant, he tried to shake my hand and I said, “This is interesting, thank you.” And I started to walk towards my car and Steve ambled laboriously behind me and as he’s walking, said, “Do you smoke?” and I said, “No, I don’t”.

“So do you mind if I do?”

And I thought, “At this point, am I going to be able to stop him?”

So he pulls out of his pocket this thing, this giant thing that didn’t quite look like a cigarette and didn’t quite look like a cigar; it was in fact a giant joint a roadside flare of weed. And out in the middle of everybody he lit this thing up and asked me if I wanted to take a puff. Okay. So obviously this was a terrible date but it’s one of just many many terrible dates and after each I would go home; I would call my mom, I would call my sister, and tell them what an awful time that I had had, and they said every single time, “Stop complaining! You’re just being too picky.”

And I thought,”That’s ridiculous”. Right, I’m gonna start showing them empirically that these are really terrible dates. So I’m going to do three things. I agreed to continue with my grandmother’s advice to date everybody until ‘least expect’ my way into true love, but I did that with some parameters. I would only meet men at bars that I knew had wi-fi. I carried a giant bag with me and my laptop inside where, once we were at the date they would invariably go terribly terribly wrong. I would pull out an email template that I had created in advance where I had different data points that I would track so that, when the date went bad I could show with empirical evidence and quantitatively why this entire thing was a ridiculous exercise.

So I would send out these email templates and I was tracking things like the number of times the guy tried to high-five me. If you want to touch me, by all means, touch me, but don’t force me to stick my hand in the air. And the number of times he abused the English language and over time I had amassed quite a bit of data. And that allowed me to make some correlations. For example — For example, the number of times I got high fived the more times that guy was to abuse the English language. The more shots he ordered, the more likely he was to lie about his job.

And speaking of alcohol, for some reason, and I still can’t figure out why, Scotch drinkers were more likely to immediately talk about kinky sex right when we sat down, than anybody else. And this may not be surprising, but lawyers were 62% more likely to pull up their mobile phone and stare into it than me, and compared to some of the other people that I went out with.

Now listen, these weren’t necessarily bad guys; they were just bad for me but the problems, actually were the dating algorithms that we’ve all come to rely on those of us who are online. Now to be sure, algorithms in dating are actually not a new thing in history; we’ve had matchmakers in every culture and my culture is Jewish and we had matchmakers, too and the matchmakers for us would be looking at things like whether or not the girl and the boy would get along, what the rabbi would say, whether the community would agree, whether or not they’d have kids and even I, in the process of setting up my friends and coworkers when asked had sort of my own formula that I was using.

Would they have the same interests? Would they get along? And what was the probability that this entire thing was gonna become a pain in my ass for which I’d be paying later on. So I want to fast forward to the worst date ever and I’m gonna spare you the details but let’s summarize by saying I smoked an entire pack of cigarettes and went through a whole bottle of wine in a short amount of time.

And then in my drunken state called my sister and I said, “That’s it. I’m done. I’m finished with online dating sites.”

And she said to me, “Don’t do that. Don’t you remember Mary Poppins?”

I said, “What are you talking about? Mary Poppins?”

And she said, “Remember in the movie how the kids had gone through all of those nannies and none of them worked and what did they do?” Does anybody remember? They made a list. Right? Those two kids started writing down every single possible thing they could want in a nanny and once they had that list it wound up going up a chimney and poof! Mary Poppins appeared. And I thought, “That’s it, Hillary You’ve nailed it. I’m going to stop “least expecting” all of this. I’m going to create my own Mary Poppins list to find a husband.”

And so in my drunken state that’s exactly what I did. I started writing. And I wrote down every single possible thing that I could think of, from height to the amount of body hair that I found acceptable to the kind of musicals that I would agree to listen to and I wrote and I wrote and I wrote. And by the end of this exercise I had come up with 72 attributes — things that I was going to demand in any person that I would date from here on out.

Now, 72 seems like a lot and I had a whole bunch of scrawled notes and it was really hard to, sort of figure out you know, who would meet all of those qualifications. I had to parse that list. So I thought about all of the people that I’d ever dated and what things on that list, the good things that they had in common. I thought about the attributes that were important to my family because my family and I are very close and there are lots and lots of them and the last thing I need is any more complaining. And attributes that were important to me.

And once I had all of that figured out I came up with a list of weighted scores so I had my top tier list and my secondary list. And in my top tier list were points that were pretty high, I had lots of different things. So I wanted someone who was Jew-ish, like me. Not religious but very culturally tied, everything. I also wanted somebody who would want to have kids with me so these are pretty normal things.

But I also thought that the stuff on my second tier list was just as important. So I weighted those slightly differently and I was very specific about what I wanted so I was looking for somebody, for example, who liked to travel but not cruise-ship travel that’s not what I would call traveling. I wanted somebody who was ready to strap on a backpack and get out and hike around. I was also looking for somebody who was not fat and not skinny, but would always weight 20 pounds more than I did — regardless of what I weighed at that moment.

So, anyhow — I had my top tier and my second tier and now that I had all of this figured out still in the same night in my drunken state, I developed a scoring system so for 700 points, I would agree to email the guy for 900 points I would go on a date and for 1500 points I would consider him for a long-term relationship. So I had set some minimum thresholds and I thought, “This is really, really smart.” Right? I am no longer gonna go out with everybody. I’m gonna have to only go out with men who meet these different criteria.

And so when I went back in I found Eric1971. He looked pretty good. I liked the way that his profile sounded, I liked the things that he wrote and as I started scoring him on a scrolled matrix I found that too many of the things from my list fell below the middle line and so as a result of that I saved myself a terrible date. Right, and having to send out a bunch of emails.

Then I found Jewishdoc57 which, I have to tell you, he was really good looking, but he also specifically in his profile said that he liked to travel and not cruise-ship travel, and I thought, “I’ve hit the jackpot”. My system is definitely working. I put him on a matrix and everything was above that middle line. I had created my own sort of algorithm and my own way to personalize these online dating websites so that I could use them as databases for my own individual tastes and needs. There’s one problem with that.

I realized that if I thought Jewishdoc57 was so amazing there were probably some people out there who did too. So, still in my drunken state I decided to take a look. And that’s when I found Smileygirl1978. SmileyGirl was very short and very thin and said that she was silly, nice and friendly and has a “genuine” sense of humor. Who is this horrible woman? And why can’t she spell? Well, SmileyGirl picked my interest, so I went deeper in and found that I had some problems because while all of these women looked pretty much and sounded pretty much the same, when comparing them to the photos that I had posted, I saw that perhaps I had a problem. And that my perfect 1500 point man may not think that I was so perfect back.

I also looked at the way that they were describing themselves so whereas I had my entire resumé posted they just used aspirational language and sounded like they were really fun to hang out with. At that moment I knew that I needed to be fearless. It was time to join JDate as a man!

True story! So I ran a little experiment. I created 10 male profiles that each fit the archetype of the perfect point man that I had developed And when I say, “Created 10 user profiles”, I didn’t just go in and type in some stuff the way that I did the first time around. I created 10 separate characters with huge amounts of data associated with each. I knew what their favorite foods were, I knew that one was in a battle with his sister and they didn’t get along because of something that happened in a car at a family vacation. I knew these men inside and out and they all scored the minimum threshold of points for me to date them. And my goal here with these 10 men was to learn.

I wanted to find out why the women who were popular on these dating websites — because the algorithm certainly wasn’t helping me, right? I wanted to know why they were popular. So I looked at their vocabulary and language. I looked at the length of the profiles; they had their senses of humor, how they describe their career, I looked at their height and their weight, what they were listing. I looked at photos, and I was also categorizing the amount of time that it took for them to make the first interactions. My family always said I couldn’t possibly email or talk to a guy first, that I would seem too aggressive. That I had to wait for them to come to me and I wanted to find out if that was true.

So started collecting data over that month and I was looking at both qualitative and quantitative data. Qualitative data show that aspirational language like “love” and “like” and “fun” would never describe something like a building as “fun.” But I learn that the popular women do.

I was also looking at quantitative data. Now we all think that everybody lies about their weight. Would you believe that more women online are actually lying about their height, rounding significantly down, than are lying about their weight? It’s one of the interesting things that I found out. After this month I had enough data now to create what I called a super profile. It was an amalgam of what I had found from the popular women that was very much personalized to me. And that’s when I gamed the system.

So I created a new profile with different photos. I was honest about my height and my weight. I can’t help the fact that I’m 5’6″ and not a size 2. But I did shorten what I had originally written in my profile significantly. I now used aspirational language, I mentioned my career but it wasn’t the focus. And suddenly I was the most popular woman on all of these websites. I had a full inbox for the first time in my entire life. I had men coming at me from every direction, every single guy out there wanted to date me. And what that allowed me to do, was to go back in and from this enormous pool of men apply my own framework, my own algorithm if you will, and the problem in doing this was that even the best of the men that were out there only scored 650 points.

My friends and family said, “Are you nuts? Everybody out there wants to go out with you. You’re still being too damn picky!” And that’s when I found this guy. So, immediately I like the way that he looked. I’m into baldies. I really loved what he wrote as his job — he said that he was an arctic baby seal hunter — and I thought he was pretty awesome. And, since he immediately scored 800 points I thought – you know – maybe my system is working.

So we started chatting, I emailed him first. My data showed me that I was able to do that, unlike what my grandmother had said my entire life, it was OK to be a little aggressive. I waited 20 to 22 hours until our next interaction although it pained me greatly every single time. I didn’t get specific about my job, the specific things that I like to watch, or the specific places that I like to go to, until our third interaction.

In the meantime I kept scoring him. Three weeks later we went out on our first date. It lasted 14 hours and it was one of the most amazing dates of my entire life. At the end of that date, I went back home, and I scored him again. He went over the 1000 point threshold and I thought, “You know what? This entire time that I’ve been listening to everybody’s advice. I haven’t been picky enough! I was too afraid to go out there and ask for and demand what I really wanted.”

Well, a year after that first date we were traveling in Petra, Jordan. Real traveling, when he got down on his knee, and asked me to marry him. We called my parents, they were thrilled, and a year after that we were married. A year after that our daughter, Petra, was born. So obviously I’m very happy and having a fabulous life but what does this mean for all of you?

Well, what this means, whether or not you’re single, is that you have to be fearless, and if everybody around you is telling you that you shouldn’t be asking for the things that you really want, tell them to take a hike. It’s OK to be fearless and to ask for what you really want and to demand it. And whether or not you’re dating or doing something else, develop your own set of data points and your own framework. Construct one for qualitative and quantitative analysis and start gaming your own system. So I wrote a book about all of this, and it’s called “Data, a love story.” It’s on pre-order now. It will be in bookstores all over the place Jan, 31st. I hope you want to read it and to tell everybody else to read it. And that’s it, thank you very much.

Further ReadingThe Secret To Understanding Male Attraction Triggers