Imagine a day in the future when you or an algorithm determine who you should date based on what sites you visit and what sites your prospective dates visit. You wouldn’t even need to fill out a 500 question compatibility test. Would you really want to date someone that actually did that, anyways?
This is the lead in for a recent Root Blog post. The CEO of Root Markets, Seth Goldstein, then goes on to coind a new phrase, “clickstream dating.”
Root Markets, or Root.net, lets people “recycle, refine and optimize (their) attention,� Goldstein said. That means that by downloading Root Markets software, people can keep an online “vault� of their Internet footprints, from the point of logging on, up until their last click.
Root’s software tracks behavior and provides a picture of personal habits in the form of digital visualizations. People will also be able to share profiles with friends to track overlapping interests, or attract new mates, in what Goldstein called “click stream dating.�
What Seth doesn’t seem to take into consideration is that dating sites already do this to some degree. While I’m the first to admin that the rudimentary matching some sites perform based on a few database fields is not the same as mining a long-tail clickstream, the idea is by no means new.
I’m fascinated by the idea of leveraging anonymous clickstreams to find matches on dating sites, or social and business networking for that matter. I’ve been writing about behavioral targeting on dating sites for a long time, although in the context of advertising. What better marketing profiles is there than a personal ad?
Here’s how I would implement clickstream-augmented dating.
Finding commonalities between the websites and blogs people read is easy. Almost too easy.
I would instead initially, limit the clickstream to dating sites. I doubt that I’m going to meet many women who visit the same websites as I do, and we’re assuming that similarities attract, not opposite.
There is a whole new type of matching engine that will have to be created on top of the raw clickstream data. Here is where I would add something like Icosystem’s proprietary Hunch Engine, as a starting point, after it’s been loaded with a base clickstream algorithm.
We can go from looking at static profiles to a real-time dynamic display of the relationships between an ecosystem of people and yourself, with dials to weight certain attributes more than others and the ability to zoom in on the important stuff using profile clustering.
Last year I threw a bunch of dating sites, founders, and VC into Prefuse. You can see it here. Take the output from the Hunch Engine and display it with a nice graphing engine like Prefuse. Make it fun to explore, without it looking like an Xbox game, er wait a minute.
Root Valult has an enormous amount of security, safety and PR management to accomplish before any of this really takes off past the 250 early adopters that even know what the “Attention Economy” means. This is going to be interesting to watch unfold. It took the digerati a few years to get behind metadata and tagging, I think general people and the media will catch on to the Attention Economy much faster. It’s easier to grok outright and basically targeted towards marketers at this point. There’s a lot of money to be made on this, right from the get-go, which is not always the case.
Technorati Tags: clickstream+dating, attention+ economy, root+markets
No related posts.
{ 9 comments… read them below or add one }
I built this a few months ago. I just never released it because i couldn’t find a way to get high user participation. Its one of those fringe things like keywords/tagging/interests searches that only 2-5% use.
If you are implementing this based on the clickstream data of members on your own site only, what additional participation is required? Isn’t the use of the site by members enough?
The clickstream would be stored at Root Valut, some other silo or the users hard drive. There is an API for the clickstream and developers are encouraged to build services that mine the data and make use of it. Somehow you would need to be able to filter the matching and tracking on a site by site basis. Perhaps an agreed upon additional tag or identfy the user as one of a dating site.
I would make users enter the clicksteam data.
I’ve got some more ideas, but implimenting it and making it useful are 2 different things.
Hey, nice article. Take a look at my dating services blog!
Although this is an interesting concept-I”m hardly basing my dating decisions on what websites someone checks out!
I think it’s a good idea and shows promise, but this is still in it’s infancy. As with the Attention Economy, Dan Gilmor has been pounding that drum hard for a long time. It’ll take years and implementations that have yet to be invented.
How long before the service is abused. One should be wary of any service that tracks ones web surfing. For all I know, this is someting that spywares have been created to do. And we all know how accepted spyware is.
My answer is no. I wouldn’t date a person that visits the same sites as I do. Never in my life. Guess why. I want to have a better person as a partner.