How To Hack Online Dating And Use Data To Find True Love

Online dating is weird as hell. You'd think this wouldn't be the case. After all, the algorithms that connect people on dating sites aren't theoretically all that different from the ones that power search engines and generate billions in revenue. So why is online dating still such a thoroughly imperfect experience? 

Amy Webb, like so many others, learned just how flawed the science of online dating is by going on a series of comically awkward dates with some pretty unbelievable characters. In her book, Data, A Love Story: How I Gamed Online Dating to Meet My Match, the digital media consultant and former journalist outlines how she "reverse engineered" online dating, reevaluated her strategy and met her future husband. 

This being the week of Valentine's Day and all, we thought it would be an opportune time to talk with Webb about her process and share some of the lessons she learned with you, oh lonely denizens of the Internet. Trust me, this is way more interesting than the romance-themed infographics we've been getting pitched all week.   

Decide What You're Looking For

Most people approach online dating like they approach joining any other social network: Set up a profile, upload a few photos that happen to be sitting on your hard drive, and fill out some personal info. The difference, of course, is that dating sites have an objective far more specific than an aimless timesuck like Facebook. 

Webb kicked off her digital quest for a mate by listing 72 traits she wanted her future partner to possess, which is how she recommends online dating newbies get started. Her initial brainstorm included everything from personal habits and marital history to work ethic and Mac vs. PC preference (Hint: John Hodgman would not have made the cut). 

The exact number of traits isn't all that crucial, as long as the list is as exhaustive as possible. "The most important thing is to sit down and write out a list," says Webb. 

Rank Your Top Traits By Priority 

Next, she broke her 72-point list into tiers: Using a combination of personal preferences and past experience, she narrowed down the ten most important characteristics and listed them as "top tier" traits. These were, in her book, the absolute deal-breakers. They were then ranked 1-10 in terms of importance. From there, she chose another 15 "second tier" traits. Not deal breakers, but still very important.

"I was looking for patterns to analyze," says Webb. "For example, there was a lot of crossover in my list when it came to family, religion and attitude towards work. When I noticed a pattern, I tried to distill from it the most important aspect of that data point."  

This approach helped her sort and tag her list and ultimately rank everything by priority. 

Come Up With A Grading System

Once her tiered list was complete, Webb assigned a total number of points to each item. Her top-tier traits were each given a total potential score: 100 for her top trait (intelligence) and 91 for her least-critical item in her top ten list (No history of cheating). The second-tier traits were all assigned a weight of 50 or fewer points, depending on their overall importance to her. 

The result was a 1000-point scale that would allow her to grade - and subsequently reevaluate - the men she dated based on the most meaningful data points. She set a threshold: Based on her initial online interactions, nobody scoring below a 700 would be worthy of an in-person date.  

It all might sound a bit obsessive, but as Webb quickly realized, this formula could have saved her the agony of going on karaoke dates with high-fiving cheapskates just a few weeks prior. 

Size Up The Competition

This is where things get interesting. Before setting up her new profile, Webb decided to evaluate the competition. The way most online dating sites work, there's no easy way to do this from your own account. So using her 1000-point grading system, Webb created two profiles of imaginary dream guys. That's right: She masqueraded not just as a man, but as multiple men, in order to see what kinds of ladies would be vying for the attention of the men she would find most desirable. In the end, she had created 10 fake profiles and interacted with 96 different women. 

This wasn't just an exercise in digital creepiness. It was a data-mining experiment of enormous value. During this phase of her quest, Webb unlocked insight into many aspects of the online dating universe, some of them more predictable than others. What correlations exist between profile popularity and hair color? What about the vocabulary used in people's profiles? How much did successful online daters refer to their career goals? What kind of photos performed the best? 

Using spreadsheets, TextWrangler and "other kludged-together applications," Webb analyzed all of this and more, manually collecting data as she went. She paid attention to things like which gender initiated conversations most and made data visualizations of the most commonly used words in the profiles of popular women. 

"If I was in another setting – like a bar, or party or work – and found someone attractive, I'd immediately look around at my competition," Webb explains. While some sites do allow you to take a look at the competition (which would save you most of the trouble here in step four), JDate does not, so Webb had to game the system in order to see the big picture. 

In the process, she discovered that the LinkedIn-esque approach she had used to build her original profile was way off-base. 

Build Your Data-Fueled Super-Profile

Armed with these new insights, Webb set out to create what she calls her "super-profile." It was concise, used positive language and wasn't as fixated on work. She focused more on her desire to travel the world than on her HTML chops. She also realized that the photos she happened to have on her laptop weren't cutting it, so she uploaded new ones based on everything she had learned by looking at the more popular users.

She showed a little more skin and scaled back on attempts at humor, which can often get lost in the context of a social profile. In short, she optimized her profile, not unlike a product page - an analogy that isn't far off from how Webb encourages people to think about their presence on Match.com or OKCupid. 

"We SEO websites all the time to ensure that they get seen first in the vast catalogue that is Google," Webb says. "Why should online dating be any different?"