It's rather funny how Stewart Butterfield keeps trying to build games, and ending up with social apps.

His first company, Ludicorp, tried to build an MMORPG called Game Neverending, which never launched, and they pivoted to create Flickr, apparently based on the same platform (which is why ".gne" URLs proliferated Flickr originally).

With Tiny Speck, he created Glitch, another MMORPG, which closed a little more than a year after launch. Tiny Speck then pivoted to launch Slack, based on a chat app they had built internally while developing their game.

It makes sense, in a weird way. MMO's are such huge projects that developers are forced to create supporting subsystems to complete it (e.g. Image/asset processing and storage system - Flickr, in-game chat and collaboration system - Slack).

I'm impressed that Stewart Butterfield is able to look at these systems and recognize which ones could be a viable (and successful!) stand-alone product.

I'm not sure why that would be all that relevant though.

If they wanted to implement a chat feature into a game it might be easier to embed this, though I could be wrong? Or if they just want a tweakable and reusable chat service to be used in-house maybe?

With even the same argument of them building a HipChat or Flowdock clone.