A few weeks, Tom Foremski at SiliconValleyWatcher complained about how dull the subject of enterprise IT has become. John Gallant, editorial director of Network World, who is involved with the forthcoming Vortex conference in the US, disagreed with Foremski's analysis: that enterprise IT is going to see some major shifts and is therefore far from moribund. Exciting? Maybe not exciting in a "get in on the ground floor and cash out soon" sense. To be honest, it's more a bloodbath waiting to happen.
Having been researching a forthcoming piece on service-oriented architecture (SOA) for Information Professional (don't expect to find the article there just yet), I can see Foremski's point. Trying to pick apart the guff the vendors come out with, I think it's a miracle anybody can make sense of what the vendors are trying to sell. The market figures for various bits of enterprise IT are not encouraging for anyone planning to invest in a vendor in many of its sub-markets. And that is where the problem lies. Enterprise IT has been sliced and diced into progressively smaller chunks by some of the analyst companies, which each group of vendors ending up in ever more obscure product sectors.
This endless division of a large market runs counter to what the users themselves are up to. I worked out quickly that to do anything sensible on SOA you have to more or less ignore the vendors and go straight to the users, unless you find articles explaining the fine differences between enterprise service buses and integration servers compelling. Years of marketing conditioning have resulted in vendors having to have some niche that they can head, even if nobody cares much about the subdivisions. That makes the enterprise IT somewhat tedious to report and does little for the perception of vendors among users.
The desire for segmentation does not go to the heart of the enterprise's problem. Almost all big companies have a large selection of broadly incompatible systems bought over the years. They would like to glue them together with as little fuss as possible. The Web has made it possible for them. I don't believe in magic bullets, so it was a bit of a surprise to find one technology turning up again and again.
Web services is quietly transforming enterprise IT as users realise they can buy a few bits and pieces that let them access software buried on a mainframe from more or less anything fitted with a browser. Not only that, once that initial connection phase has been dealt with, they can start to bolt a layer of software onto the front that lets them package different server applications and offer them to third parties or customers. The problem for the enterprise IT vendors is that much of this work has to be done by the customer or by a service company, such as CSC or IBM. Packaged software does not fit neatly into this environment although there are gaps that the software can fill.
The emphasis is more on breaking up the packaged software into components that can be offered up to users separately, as individual services. As the users do this, they are finding that they have a lot of duplication in what they run. Gradually, they are going to reduce this additional fat and align the system around a few core components. That is not good news for vendors, unless some of them decide to live with lower licence revenues and fully the embrace the idea of turning their offerings into components and glueware in order to displace more expensive incumbents.
The trouble is that, right now, many of the software companies keep plugging away at a business model devised 20 or 30 years ago, pretending the packaged software business is alive and well. They just want to sell users something big, shiny and new, not just some glueware for the old bits. At the same time, the service companies have cottoned on to things such as open-source frameworks for providing the glue. That is cheap after all.
Foremski used the example of enterprise IT to describe how the Fortune 500 companies will fall prey to 'new rules' companies. But many of those big companies are not paralysed by their systems. They are getting hold of what it takes to update their systems. Some will inevitably make a mess of it but many others will make the transformation work. The whole Web services setup has given them a better escape route than what the same technology does for those 'new rules' entrants. The people desperately in need of the new rules are the IT vendors using a technology and marketing strategy straight out of the 1980s and 1990s where it was OK to confuse your customers because you were going to tell them what to do anyway. Now, those customers are going to be getting their own back and licence fees are firmly in their sights. Microsoft's decision to aim at the medium-sized companies for much of what it is doing looks to be a much safer bet.