In "The Starfish and the Spider: The Unstoppable Power of Leaderless Organizations" I finally read a line I've been pushing for a while about the changes being wrought by cheap high-performance computing and telecommunications (and the associated software). IANAE, but Ronald Coase collected his Nobel Memorial Prize in Economic Sciences from the Bank of Sweden for his work on the nature of the Firm. Essentially, AFAIK, it illustrated that the firm exists to reduce transaction and co-ordination costs. Back in the industrial day, that was by getting larger and forming close-packed command and control structures.

But that advantage is now available at a much smaller scale, down even to the individual, and so the compulsion to cram into a single building in order to ease organisation, despite the costs of such an architecture, is gone. Further, in a rapidly changing environment, the rigidity and inertia of large tightly bound organisations leads to a diseconomy of scale as responsiveness rather than perpetuity is rewarded.

It's still possible to be big, but you will be flat. Stovepiping up the stack from your infrastructure will not be rewarding. The shearing stress of the different rates of change, increasing as you go higher in the stack, will tear you apart. Something Microsoft, newspapers, telcos, and perhaps eventually Google, suffer from. Big infrastructure, as long as it remains just that, will persist, and so no doubt will the temptation to rise into services. Resist that temptation and all is well, fall victim to it and suffer the consequences. The most damaging consequence to an infrastructure provider entering the services market is they have made themselves a competitor of all their customers, which most customers don't like.

While I applaud the "small pieces, loosely joined" model, there is still a sweet spot for the large infrastructure utility, question is, how far down the stack do you need to go to find it?

Check out the book, while it is overly gushy, and ignores a more nuanced investigation of some of the claims, and provides one of the worst analogies for copying I've read outside of the recording industry's propaganda, there is much useful in it.

Large, vertically integrated information organisations are suffering, this book is one illustration and attempt at explanation, for a more oblique example, listen to this and think of Telecom.