Here is Celent writing about data and coordination, which I think boils down to the idea that IT is gradually becoming a core skill of other business functinos.
I think that’s right. I also think that IT progress can be categorized.
There’s innovation (new apps), which largely supplants business models. All very exciting but, realistically, it’s the least important part of technological innovation.
There’s computing power improvements, which are interesting, but I think that once we’ve gotten over the ability to create and manipulate text and images instantaneously, most businesses find they’re bound by…
Organizational Innovation. Over to David Leonhardt, interviewing Alexander Field about the productivity explosion during the Great Depression:
Organizational innovation also played a role. In railroads, treaties now allowed unlimited freight interchange. Rolling stock — railroad cars — from one road could move onto tracks owned by another, and while there, discharge and pick up cargo, and even be repaired in a “foreign” yard. The agreements and uniform tariff schedules that permitted this were critical in enabling U.S. railroads to carry more freight and almost as many passengers in 1941 as they did in 1929, using many fewer employees, cars, and locomotives.
Organization innovation is all about finding new ways of executing old ideas. This is what Celent is talking about.
Standards problems are common; look at Japan’s power grid, for goodness sake. That power grid problem seems like madness now, but any time there is a new technology, its implementation is going to be fragmented.
The insurance industry is not immune to this. This is a business where most companies effectively raise capital every year (in annual reinsurance contracts) but don’t have a standard way of transmitting largely homogeneous data (insurance limits, premiums, class of business, etc). Data problems are a big deal.
I think that this will eventually change, but probably not until it becomes much easier to radically change the schema of a database after it’s been populated. That’s a pretty tough trick to pull off at the moment.
Not Unreasonable 
