Microsoft Mixer

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

~ the obituary ~

mixer was microsoft's third attempt to own a part of pc gaming culture it had repeatedly lost. xbox live owned consoles, but the desktop streaming layer had been twitch's since 2014, and microsoft was watching the next generation of gamers grow up there. so in august 2016 microsoft bought a small seattle startup called beam, rebranded it as mixer in 2017, and tried to buy a winning streaming service the way you would buy a server cluster.

i never streamed on mixer. by 2019 i was a working engineer in turkey doing edge ai on uavs, and the part of the day i was watching games on a screen at all was twitch on a second monitor while a model trained. mixer was an english-language us-eu streaming service for the kind of pro broadcaster i was not. the closest i came was reading the verge about the ninja contract.

the deal was the headline. tyler "ninja" blevins signed an exclusive at mixer in august 2019 for what was reported as $20-30 million across multiple years. michael "shroud" grzesiek signed a similar deal in october. neither of them brought their twitch audience with them, because nobody installs a second streaming app for one creator if they already follow ten on the original.

~ the case file ~

~ the climb ~

beam launched on january 5, 2016 in seattle, founded by matthew salsamendi and james boehm. its pitch was sub-second latency between streamer and viewer, faster than twitch could do, plus an interactive overlay system that let chat vote on what the streamer would do next. it was a real engineering improvement on twitch. microsoft acquired beam in august 2016 for an undisclosed amount and rolled it into the xbox group.

the rebrand to mixer happened in may 2017. the integration with xbox live and windows 10 went live the same year. by 2018 mixer had a real product and a real roster of streamers, and microsoft had pulled some xbox-adjacent talent across to broadcast on it. it was small at around 1-2% of twitch's hours-watched, but growing.

~ the slide ~

streaming is a network-effect business. you watch the streamer your favorite streamer watches. the way to grow a streaming platform is to bring over the social graph, not just the talent at the top of it. microsoft tried to skip the social graph by paying the talent at the top of it, which was the part that was supposed to bring everybody else along.

in q3 2019, the quarter after the ninja signing, mixer's hours-watched actually fell 11.7% even as content hours rose 188%. that means more streamers showed up to broadcast and fewer people showed up to watch. the streamlabs and newzoo joint reports confirmed the number. ninja himself reported that his viewership on mixer was a fraction of his twitch peak. the deals were paying for the ceiling and the floor was empty.

~ the actual end ~

on june 22, 2020 microsoft announced mixer would close on july 22, the same day it announced a partnership with facebook gaming. existing mixer streamers got migration tools and an option to be co-streamed onto facebook. the contracts with ninja and shroud were terminated, with both reportedly keeping a portion of the pre-paid contract. ninja returned to twitch within a year. shroud also returned to twitch within months.

microsoft had spent the better part of $50 million signing two streamers and several hundred million on platform infrastructure to learn that streaming markets are not the same shape as console-platform markets. xbox live had gotten there by being on the box you bought to play games on. mixer was an app you had to install on a second device for a service you had not used before, hosted by the people you had to be told to follow. the second of those is harder to buy.

@}-->--

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

« previous  ·  random  ·  next »