Data and dating: Amy Webb gives her thoughts on online romance

http://www.guardian.co.uk/commentisfree/2013/jan/29/data-dating-amy-webb-online-romance

Version 0 of 1.

Data and dating expert Amy Webb joined the Guardian for a live Q&A on online romance – and we've collected the best of it below.

Of all the data you collected, what was the most revealing? And essential?

The most important thing I learned was that online dating should be treated for what it really is: a big database. Since you can't control the quality of information being added by other people, it's on you to create a system to parse it. So the most essential lesson was this: You need to define, with great detail, your exact target audience, and then market yourself – you, the product – to land a committed relationship with that audience.

One surprising element was profile length. I thought I should explain everything I do and like. But we've all met "over-sharers." The people who tell you their entire life story within the first five minutes. You wouldn't over-share at a party, right? Then don't do it online. My data showed that succinct, pithy profiles do best.

You've said that we create "idealized versions of ourselves" on our online dating profiles – do you think we're capable of creating useful dating data?

The problem with online dating is that algorithms use the information we enter ourselves. Algorithms don't work. Bad data in = bad data out. Think, for a moment, about the people who "smoke only when they drink." I'd call them smokers. We tend to enter information on dating sites, in Facebook, in Twitter and elsewhere, aspirationally rather than objectively.

What's the best way to tell if someone is stretching the truth on their dating profile?

It's not always easy to see if someone's stretching the truth when you meet them online. I recommend talking to someone on the phone a few times, emailing back and forth, before meeting them in person. If you're tech-minded, you can also reverse-trace IP addresses, look up usernames via knowem.com. Do a reverse image search using TinEye.com, that sort of thing. I did extensive background checks on @drwoolf before we went out in person. If you have something embarrassing or shameful to reveal, certainly don't mention it in your profile. Pick a good time IRL.

What are some ways online dating sites can make navigating the data deluge easier for users?

To make the process more user-friendly, dating sites should clean up their interfaces to start. Dating sites should court users little by little, just as we court our partners in real life. Don't ask us to answer 20 questions right away. Don't submit us to a Myers-Briggs style questionnaire. Matching algorithms can't work as is. Work to solve for user-generated data problems, which is a hard nut to crack.

What's one piece of advice would you give someone new to online dating, who isn't a data guru?

Well, I make the point in my book that I'm not a dating coach, or a relationship guru. I'm just a woman who was tired of going out on bad date after bad date. And I thought I deserved better. Keep in mind that if you're feeling rejected or that you're going on dozens of awful dates, it's because digital media has accelerated what is usually a longer process. We tend not to date 12 new people in a single week. Yet we can easily meet, chat with, initially feel attracted to and then horrified by 12 new people a day because of online dating. It's (most likely) not you. It's the system. My advice: game it on your own way. Figure out exactly what you want and then market yourself to get what you're looking for.

You put a lot of time and effort into your online dating profile! How do we keep data-tracking manageable?

Well, I put a ridiculous amount of time into most things. Usually to my own detriment. I'm a big believer in self-quantification and data tracking in general. Keep a journal, a spreadsheet, or build yourself an app. The point is to create some kind of system. Most people don't go to the grocery store to buy "stuff." Your list probably doesn't say: produce. Dairy. Paper Products. Meat. You don't need to obsess like I did, or to track all that data. But do write down some kind of framework.

What comes next after Data Love Story? What kind of data do you find most interesting now?

So this is somewhat embarrassing to admit, but Brian and I went a little overboard tracking metrics on our daughter. We were tracking all possible data. It went way beyond poop. We were trying to figure out when she was most attentive so that we could occupy her - flash cards, me playing piano with her. By her six month visit, we asked her doctor to give her a grade, so that we knew whether or not we were making good progress. He gave her an A, and he gave us a C-. He told us to put the binder away and to stop making scatter plots of her … well, of her everything.