I think I've touched on this question before but with the last technology as infrastructure post it seems like another opportunity to talk about the intersections between this topic--thinking about technology as infrastructure--and about the sort of small scale/cooperative economics that I was writing a lot about a couple of months back.

The question on my mind at the moment is, "What do the business models of technology firms look like, in a software-freedom-loving, non-corporate/cooperative-business way?"

And I'm not sure what the answers to this question are. Not really. I've been thinking about business models for the producers of software/technology services during earlier moments.

We have the example of the 70s and 80s when the prevailing technology companies were ATT and IBM. ATT made their money selling phone service, and licensing UNIX. IBM made their money selling mainframes. In the eighties and nineties we had the prevailing Microsoft lead "proprietary software licensing" business models, where consumers paid for the legal write to run code on their computers.

In the nineties and early naughties the successful business models were either from people buying hardware (ie. Sun Microsystems and IBM) or people buying support for operating systems (ie. RedHat). We've also seen some more stable business models centered around subscriptions-for-services (this seems to be what all the successful startups are doing), and more of the time honored selling hardware, and there are some support-services based companies that remain successful (eg. RedHat), the support market consolidated a lot recently. And it's not like the Microsoft-consumer model doesn't still exist.

So when we look at "infrastructural technology" it sure looks like there are some kinds of businesses that will continue to flourish:

  • The people making mainframes/servers and the high level computing systems that provide the infrastructure.
  • The people who provide the tools that make low level tools successful and useful to users. (eg. What UbuntuOne provides on top of SSH and rsync; What gmail provides on top of IMAP; What MobileMe provides ontop of WebDAV/CalDAV/IMAP).

These strike me as rather conventional business models, given the history. Does infrastructural computing also:

  • further the development of subscription-based businesses?
  • create a new kind of challenge in customizing solutions for organizations and groups that translate raw resources into "finished output?" Is this too much like IMAP --> Gmail?
  • [other possibilities created by you, here]

I'm trying to approach this by asking myself "what creates value in this market," rather than "where's the money." It strikes me that value exists in making systems "work" in a way that's customized to the task at hand. It strikes me that value is created when individuals and organizations are able to take ownership of their own data and computing. Gmail is valuable, but running my own IMAP server is more valuable. Running my own IMAP server without the fuss of needing to personally manage the hardware and software of the server is even more valuable.

What else does the hive mind have for us?