ScientificMatch.com

» see scientificmatch.com as it lived, on the wayback machine

~ the eulogy ~

Two thousand dollars settles the question of whether I was ever going to be a customer here. That was the entry fee, $1,995, not negotiable, for a cheek swab and a login. ScientificMatch was built for Americans with disposable income and a particular kind of faith in laboratories, and I was outside both halves of that.

Eric Holzle launched it in December 2007. He was a mechanical engineer rather than a geneticist, and the thing he was selling came out of the immunology literature: the major histocompatibility complex, the gene cluster that determines what your immune system is able to recognize. Work going back to the mid nineties found people tend to prefer the body odor of partners whose MHC genes are unlike their own, and the usual explanation is that the mixture gives children broader immune coverage.

So the pitch was that attraction has a measurable component and a lab can measure it more reliably than a room full of people can. You paid, they mailed you a swab, you scraped the inside of your cheek and sent it back, they read six immunity markers, and the site then showed you only members whose markers were sufficiently different from yours. Everything else you typed in yourself, the same as on any other dating site.

The press coverage was enormous and almost entirely amused. The Economist ran it in January 2008 under a headline about the scent of a woman. NPR and Bloomberg did versions of the same story within about ten weeks, and so did most of the American papers.

~ the dossier ~

~ the unraveling ~

The science was thinner than the marketing. The t-shirt studies were small, run under laboratory conditions, and what they measured was rated pleasantness of smell, which is a long way from a marriage. The original work also found the preference reversed in women taking oral contraceptives, a result that never made it onto the sales page. Later research put the correlation lower than the first papers suggested.

The bigger problem was arithmetic. A dating service is only as good as the number of people inside it, and a two thousand dollar door charge guarantees that number stays small. Then add a filter that removes everyone whose immune markers are too close to yours and you are looking at a very short list in a very large country. The premium price was supposed to signal seriousness. What it actually signaled was that there was nobody in the room.

~ what it became ~

The site was dissolved in early 2012 with no farewell and no press release, which is how most dating startups end. Its Wikipedia article is a stub, and the only See Also link on it points at Mouse Mingle, a dating site for adult Disney fans, which is a fair summary of where novelty matchmaking sits in the taxonomy.

The idea keeps coming back, attached to whichever consumer DNA company is having its moment, and it keeps hitting the same wall. The effect is small, and people choosing partners are not optimizing for their hypothetical children's immune systems. ScientificMatch was early enough that the coverage was curious rather than skeptical. That window has closed.

@}-->--

              )
             (_)
            .-'-.
            |   |
            |   |
            |   |
            |   |
          __|   |__   .-.
       .-'  |   |  `-:   :
      :     `---'     :-'
       `-._       _.-'
           '""""""

« previous  ·  random  ·  next »