This Is My Jam

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

~ the eulogy ~

this is my jam launched in september 2011, built by matthew ogle and hannah donovan after both left the echo nest about six months before spotify acquired it. the conceit was one constraint, posted on the about page in a single sentence. you pick one song, that song is your "jam" for seven days, your profile shows that one track and nothing else. people followed each other and the front feed scrolled past a steady stream of single-jam picks. once a week your old jam dropped off and a new one took its place.

the engineering underneath was a thin shell over embedding apis. the audio came from youtube and soundcloud and hype machine, plus a handful of smaller embed sources. this is my jam never hosted a file. that kept costs down and let two founders run the whole site for years on a small budget. by the time it shut down, over two million tracks had been picked as somebody's jam.

i remember reading about this is my jam on twitter at the time, around 2012 or 2013, and not signing up. by then i was in my late twenties, the music-sharing site that mattered to me was last.fm scrobble data flowing in from foobar2000 and a turkish forum thread about new artists. a site that asked me to post one song a week as a public statement was the wrong shape of social object for my use, and probably the wrong age bracket for the user base it ended up with, which read younger and more london-and-brooklyn coded than ankara-and-istanbul.

~ the dossier ~

~ what it actually was ~

the single-track constraint did most of the work the site was famous for. on a normal music network the question is "what are you listening to". on this is my jam the question was "what is your one song right now". the answer carried more weight because you only had one, and the next person down the feed only had one, so reading a friends page felt like a row of seven-day moods instead of a scrolling list of every recent play.

this was the opposite of the last.fm scrobble model, which was passive and total. matthew ogle had built parts of last.fm and knew what the data-firehose model could not do, which was tell you what a friend currently cared about. one song at a time was the answer.

~ how it ended ~

the closing post went up august 9 2015. matthew ogle had joined spotify in a product role on discovery, hannah donovan had become product design lead at drip. the founders said they had spent the prior year looking at options and could not find a way to keep operating. the underlying problem the post named was embed-source friction. as the streaming services tightened their embed rules and changed their apis, the cost of keeping the site working stayed flat while the user growth did not. there was no way to monetize a free site that hosted no files.

the site went to read-only on september 26 2015. every user keeps a profile page, every old jam is still embedded, the playback still works through the upstream apis as long as those apis still serve those tracks. ten years later the archive is still online at thisismyjam.com. for a 2010s indie social product, that is unusually clean death management.

@}-->--

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

« previous  ·  random  ·  next »