Monday, April 10, 2006

IBM 4 Enterprise

Redmonk's James Governor is undoubtedly being mischievious mischievous in his latest post - The Definitive Proof that Ruby on Rails is enterprise technology.

Here is his so-called proof.
  1. A recent recruit at IBM is doing things with Ruby.
  2. Presumably IBM hired this person for his Ruby skills.
  3. Therefore IBM has endorsed Ruby.
  4. Anything IBM does is enterprise.
  5. Therefore Ruby is enterprise.
But this argument overlooks a critical decoupling that is evident in IBM's history - between R&D and Sales.

The history of IBM certainly includes some pretty impressive research, including a number of Nobel Prizes. IBM employed Mandelbrot. IBM made a computer (Deep Blue) that can beat grandmasters at chess. IBM also takes credit for the invention of the relational database and the personal computer - inventions which were (much later) widely adopted by the enterprise.

IBM also has a history of trust, and a powerful sales organization. Nobody gets fired for buying IBM, went the whispers.

Sometimes there is a very long lead time between the invention and the enterprise adoption. One of the trust factors in IBM's favour is that it doesn't have a history of rushing new technology prematurely to market. IBM may have invented the relational database, but it didn't produce a commercial RDBMS until long after the first movers.

From many points of view, a disconnect between Research and Sales would be regarded as an organizational weakness. But it is also a strength, and reflects the fact that Research and Sales have different loyalties. Enterprise customers can trust the IBM salesman, the argument goes, precisely because he is closer to the enterprise.

IBM has changed a lot in recent years, and perhaps the historical disconnect between Research and Sales is now less evident. But the idea that an experimental use of a technology within IBM is equivalent to enterprise endorsement is still far-fetched. One swallow doesn't make a summer.

Technorati Tags:

No comments: