Mate search, one of the chief activities of humans, interacts in interesting ways with modern technology. Computers can gather and analyze enormous quantities of data; how best can they be used to improve on realspace search strategies? Sexual norms and behavior have changed a great deal over the past century, in part due to technological progress in contraception; has the result been to make the project of finding a mate easier or harder?
When I first signed up for okcupid 10 years ago my recent ex was my top match. Oof. I was hard core on my profile, I listed the characteristics I wanted in a husband and made it clear that I was looking for marriage and children. My friends criticized me, said I was scary away guys, I said that was the point. I didn’t want just any guy, I wanted the right guy. My now husband was a 97% match with me. He was so into my weirdo self that instead of messaging me online he stalked me, found me on Facebook and showed up to an event I rsvped to and asked me out in person. We recently had our 10 year wedding anniversary.
Interestingly I met my best friend through a guy we both matched with who friendzoned us both. He was too good looking for us. Lol But he suggested we be friends because we had so much in common, it’s kinda freaky. We even had kids only months apart from each other.
I don't think your strategy applies to Tinder-style apps – which currently dominate the market. A male on such apps can only message the women who have swiped right, which is typically a small enough group that filtering is not an issue. The goal for normal males on such apps is thus to maximize the number of interesting females who swipe right – the cost of dealing with uninteresting females who swipe right is minuscule in comparison. I think the best strategy for most males on such apps would be to get professional assistance with photographs and profile creation, and possibly also with texting, which is the second great filter before the actual date. Oddly enough, this industry doesn't appear to have taken off yet.
I built a dating app wherein users chose dealmakers & dealbreakers and they were matched using simple boolean logic. For example, if a user chose, "My match likes dogs" then the other user must have chosen, "I like dogs," and vice versa for the other user's selections. As can be easily predicted, since people tend to be picky, this does not work unless there are a lot of users on the app. We tested various tweaks such as limiting the number of selections, restricting the list to only values-based items, and so on, but it is simply very difficult to compete in the dating app market. Most apps are free and the market is saturated (historically, over 10,000 dating apps have been created with a pareto distribution for active users). We invested $15K in personal capital for marketing by targeting values-based communities which we thought might increase the probability of matches, but we never got beyond a few thousand users. Many users loved the concept, but we made a mistake of not targeting word-of-mouth communities such as colleges where such an app can spread more widely. Most large dating apps started on colleges (as did social media companies). We had chosen against it because our initial target market didn't include college students as we believed older people are more likely to know what they want from a values-based perspective (which we hypothesized maximized likelihood of dating success).
As you noted, a lot of the OkCupid-related strategies are dead because OkCupid completely changed after they were bought.
My new hypothesis for my personal dating strategy (besides apps and social circle, etc.) is to create an online profile on apps such as Instagram, etc., showing my life activities. Then potential mates can see more context to my life. My friend recently found a wife that way.
A (to my mind) huge issue with dating apps is that we are terrible at consciously knowing (let alone expressing) what we like in a mate and that, even if we were, that may not correlate as strongly as thought to likelihood of long-term relationship.
I think that either some a) psychographic approach that relies on data mining long-term married couples or b) a friends-only prediction market that pays our friends for each month the relationship goes one would be promising approaches to supplant the current lacklustre methods.
The problem I'm seeing with Tinder et al is that women tend to focus on the most popular men, who take advantage of their popularity to extract casual sex from many women, who are then fed up and don't give others a chance. Not a game I'm playing myself, but my kids are approaching the age of relationships and I worry about them.
Great post. It seems that this is a solvable problem and that whoever solves it can get rich by doing so (and while making the world a better place). I have no experience with dating apps, but it seems that the key to success is to get a positive feedback on realistic matches, not swipes or views.
Just spitballing, one way to improve matches seems like it would get people with similar tastes and interests to attend group functions with zero prior sorting for physical attractiveness. Get people to mingle and meet each other and group out from there. I would also allow feedback on people to come out of the mingling, again without regard to physical attractiveness. Cads and psychos need to be screened out quickly, but everyone else could spend time in small group settings getting to really learn if there is chemistry.
When I first signed up for okcupid 10 years ago my recent ex was my top match. Oof. I was hard core on my profile, I listed the characteristics I wanted in a husband and made it clear that I was looking for marriage and children. My friends criticized me, said I was scary away guys, I said that was the point. I didn’t want just any guy, I wanted the right guy. My now husband was a 97% match with me. He was so into my weirdo self that instead of messaging me online he stalked me, found me on Facebook and showed up to an event I rsvped to and asked me out in person. We recently had our 10 year wedding anniversary.
Interestingly I met my best friend through a guy we both matched with who friendzoned us both. He was too good looking for us. Lol But he suggested we be friends because we had so much in common, it’s kinda freaky. We even had kids only months apart from each other.
I don't think your strategy applies to Tinder-style apps – which currently dominate the market. A male on such apps can only message the women who have swiped right, which is typically a small enough group that filtering is not an issue. The goal for normal males on such apps is thus to maximize the number of interesting females who swipe right – the cost of dealing with uninteresting females who swipe right is minuscule in comparison. I think the best strategy for most males on such apps would be to get professional assistance with photographs and profile creation, and possibly also with texting, which is the second great filter before the actual date. Oddly enough, this industry doesn't appear to have taken off yet.
I built a dating app wherein users chose dealmakers & dealbreakers and they were matched using simple boolean logic. For example, if a user chose, "My match likes dogs" then the other user must have chosen, "I like dogs," and vice versa for the other user's selections. As can be easily predicted, since people tend to be picky, this does not work unless there are a lot of users on the app. We tested various tweaks such as limiting the number of selections, restricting the list to only values-based items, and so on, but it is simply very difficult to compete in the dating app market. Most apps are free and the market is saturated (historically, over 10,000 dating apps have been created with a pareto distribution for active users). We invested $15K in personal capital for marketing by targeting values-based communities which we thought might increase the probability of matches, but we never got beyond a few thousand users. Many users loved the concept, but we made a mistake of not targeting word-of-mouth communities such as colleges where such an app can spread more widely. Most large dating apps started on colleges (as did social media companies). We had chosen against it because our initial target market didn't include college students as we believed older people are more likely to know what they want from a values-based perspective (which we hypothesized maximized likelihood of dating success).
As you noted, a lot of the OkCupid-related strategies are dead because OkCupid completely changed after they were bought.
My new hypothesis for my personal dating strategy (besides apps and social circle, etc.) is to create an online profile on apps such as Instagram, etc., showing my life activities. Then potential mates can see more context to my life. My friend recently found a wife that way.
A (to my mind) huge issue with dating apps is that we are terrible at consciously knowing (let alone expressing) what we like in a mate and that, even if we were, that may not correlate as strongly as thought to likelihood of long-term relationship.
I think that either some a) psychographic approach that relies on data mining long-term married couples or b) a friends-only prediction market that pays our friends for each month the relationship goes one would be promising approaches to supplant the current lacklustre methods.
The problem I'm seeing with Tinder et al is that women tend to focus on the most popular men, who take advantage of their popularity to extract casual sex from many women, who are then fed up and don't give others a chance. Not a game I'm playing myself, but my kids are approaching the age of relationships and I worry about them.
Great post. It seems that this is a solvable problem and that whoever solves it can get rich by doing so (and while making the world a better place). I have no experience with dating apps, but it seems that the key to success is to get a positive feedback on realistic matches, not swipes or views.
Just spitballing, one way to improve matches seems like it would get people with similar tastes and interests to attend group functions with zero prior sorting for physical attractiveness. Get people to mingle and meet each other and group out from there. I would also allow feedback on people to come out of the mingling, again without regard to physical attractiveness. Cads and psychos need to be screened out quickly, but everyone else could spend time in small group settings getting to really learn if there is chemistry.
Your calculations made me laugh, but very sweet how you describe your wife.
1. They have only changed in a small subset of the world, primarily white and Christian.
2. In the Amy Webb paragraph, you wrote no instead of not.