History doesn’t repeat itself, but it does rhyme

When Mark Twain said that, he almost certainly was not thinking about the fate of network operators and mini-computer companies, but it seems more and more likely that 20 years from now Verizon, AT&T, Sprint, Comcast, Cingular and T-Mobile will be where DEC, Prime, Data General, Stratus, and Tandem are today.

Sun was right, at least about one thing – the network is the computer. Now, more than ever people get IT services from the cloud, whether its Gmail or Salesforce, or a hosted IP PBX. That trend, if anything, seems to be accelerating. So, if the network is the computer, what forces will shape the architecture of that computer.

In the 1980s mini-computer companies competed with each other at an architectural level. Customers invested in a DEC architecture or a Data General architecture. Their applications ran on one or the other but not both. When the PC companies burst on the scene, the mini-computer companies reacted (in character) by building architecturally differentiated “PCs” – remember the DEC Rainbow.

The PC companies like Dell and Compaq did not try to differentiate their architecture. Instead, they worked to improve the price performance of their products while maintaining compatibility with the standard PC architecture. They also innovated at the business process level, experimenting with direct distribution, flexible supply chains and low cost, high volume manufacturing systems

As PCs became more powerful and low cost servers began to displace core minicomputer applications, the minicomputer manufacturers used their substantial R&D capabilities and their access to capital, not to compete under the new rules but to try and extend the life of the old rules. Instead of focusing on reducing costs, they regarded their current cost structures as a given and looked for ways to improve margins by selling complete “value added” solutions.

For those of us who have been around for a while, the rumblings of the telcos about moving up into the applications space, or the cable companies moving more aggressively into content, or the wireless operators focus on value added “on deck” services all start to look familiar. None of these players seem to embrace the decentralized innovation that created the PC and the Internet. None seem to be using their substantial R&D capabilities and access to capital to build the best price performance connectivity. None seem to be innovating at the process level. They all seem obsessed with innovating at the architectural level, creating supersets of the open standards that created the web, and reaching up into the applications layer to create high margin vertically integrated “value added” solutions.

There are some substantial differences between the situation today and the one the minicomputer vendors faced 20 years ago, most importantly regulation of the last mile and wireless spectrum, but there are enough similarities to appreciate the rhyme.

The question is – who will be the Dell of the network service providers? Will one of the existing players recognize that decentralized innovation on a modular, standard platform will inevitably win in the long run and take the steps to position themselves as the low cost, high quality provider of standardized network services, or will a new player capitalize on that opportunity? My bet is the latter.

Recommended in The Nature of the Internet