Wednesday, November 30, 2005

The Digital Divide in the Corporate World

Walter Wriston did pioneering work in the use of IT in the corporation, and he foresaw that it would flatten hierarchies politically as well as in the corporate world. Yet Citibank remains one of the most hierarchical organizations around, and 25 years after Wriston's pioneering work they were still doing management workshops to get people to stop hogging information to build their powerbases in the company.

That story I got from a friend whose wife worked for Citibank a few years ago. But I've had plenty of comparable experiences of my own. These cases are interesting and important, for they are absolutely material clues to the frequent failures of IT projects. Interestingly also Microsoft, arguably one of the more successful software companies around really succeeded by focusing on the easy stuff, some programming languages at first but then personal applications. That is to say they solved the problems of secretaries and other administrative staff before they ever got into hairy stuff like networking infrastructure and server software, not to mention business applications which they are now entering, and which are truly an intractable area of only a few hard fought successes. Or, to put it differently they built their fortune based on solutions which left the hard part - business analysis - to the user. And now they're getting to the hard part.

I once did an interview with Prof. Edsger Dijkstra, in which we explored in depth his insights around the frequent IT failures in business, and he was very clear that many business problems can be mathematically quite intractable, and Database logic is about the hardest thing there is.

But it is the control issue that really kills a lot of IT projects. My own most interesting experience revolved around a Decision Support technology in the shipping company where I worked in the early 80's. In conjunction with an MIT based OR consulting group we prototyped a Decision Support workstation for the company's fleet of chemical tankers, including evaluation of all combinations and permutations of a given trading pattern, and the ability to do load maximization and optimal pricing. We had good economic models indicating that the project had the potential for upto 20-30 million dollars of incremental intramarginal profits compared to the then somewhat computer supported, but really manual/personal decision making about major changes in trading patterns. One of the interesting examples was a change in voyage pattern we discovered at a certain point, yielding an extra $1 million per trip, but by the time we discovered the pattern, we had already missed the opportunity three times, in essence because it involved the need of three people talking about different parts of the world. That was the clearest ever indication of why such technology could change things for the better.

The project got shot down, essentially because the then sixty-five year old founder of the company had an emotional aversion to giving up decisions he ultimately regarded as his prerogative, and the head of operations, who perfectly well understood that I was actually right, saw the project as a threat to his job.

Then in 2001 I was in the company as a vendor, and the new CIO, who had a background at AMR Corp, and therefore understood yield maximization, told me he was then implementing the project which had been abandoned in 1987. I never found out what became of it, but I wondered how much was the compound return on an extra $25 million per year for 14 years? I never did the calculation.

In short "management" in many cases feels as much threatened by information technology as factory workers do by machines. It is this friction more than anything which is an actual manifestation of the "digital divide," in a practical way, which means that many people who do have access to the technology, misuse it for trivial tasks because they are afraid of it.

The above scenarios are just another way of looking at how we often fail to make use of the availability of IT solutions even when they are plentifully available. In this sense the digital divide is here, not there. And as long as we don't understand it here, we can't solve it there. To think that solving the digital divide means providing access to personal computers and the internet is silly nonsense. Bridging the digital divide requires extending useful applications and solutions to people in a way that's easy to use and access. It may be through an internet café, a cell-phone, a fat client or a thin client as the case may be, but success means usability, not how it's delivered.

Copyright © 2005 Rogier F. van Vlissingen. All rights reserved.

No comments: