Get over it. Great advances in software have promised great advances in productivity--whether in the realm of the Internet, middleware, or system software--but these products, just like ourselves, have to be deployed in the real world, where dreams usually don't come true.
Consider some not-so-distant history. In the' 70s, Al held great promise. Expert systems were going to help businesses automate complex decisions by encoding the experience of highly knowledgeable people. In the '80s, we were told that object-oriented programming was going to speed the implementation of complex systems by maximizing re-use. The mainframe was going to die, and open systems running on client-server architectures would make systems interoperable. While some impressive applications of expert systems and object-oriented programming were implemented, no AI and OO software vendors achieved critical mass. IBM S/390s are not only still in the picture but have grown as enterprise servers. And anyone maintaining Unix systems can tell you that interoperability comes at a cost.
How can so many smart people have been so wrong?
TECHNOLOGY ISN'T IMPLEMENTED IN A VACUUM
While C-level management recognizes that replacing an application, DBMS, or operating platform is costly, too often it overlooks the cost of integration. Since Forrester estimated that $82.5 billion was spent in handcoding interfaces in 1998 and that that figure would increase by 30% in 1999, this could be an important oversight.
Why is integration so costly? Because you use IT to manage your business, and you can't stop doing business just because you want a "do-over" in IT. While a data warehouse or e-commerce application may seem independent of operational systems, it only brings value if it shares data with the applications that run the business.
Because of the pressure to minimally disrupt the business, integration projects have often employed shortcuts so that historical data did not have to be rebuilt. The result is "secret" schemas, modifications made in such a way that they are unknown to the file system or data manager. Frequently undocumented (or else documented in a variety of places), these tricks, coupled with the sheer complexity of some operational databases and the requirement that mission-critical applications be minimally impacted, make it difficult for companies to quickly deploy and benefit from advances in software technology.
THEORIES ARE SIMPLE--REALITY IS NOT
Part of the problem is that technology evangelists are unaware of how IT really works. Look at who large companies go to for technical vision--the industry analysts. While it is true that expertise is important (who would want to be their own oncologist?), consider the economic structure of most of these firms. Vendors pay them for marketing advice, end users for purchasing advice, and both for conferences. Several are affiliated with investment banking firms that take technology firms public. That is not to say that the industry analysts are corrupt or uncaring. Nevertheless, the structure of their business places certain pressures on them that may make it hard for them to serve all masters.
Even more important is whether they have the background to give the advice they are giving. Too often they have only a conceptual understanding of the problems being solved, with little operational experience--hence, no understanding of the cost of deployment. Or, if they have operational experience, it may be limited to being an RDBMS database administrator. If they are visionaries, they may be talking about what should be (in an effort to encourage vendors) as opposed to what vendors are doing or plan to do in the foreseeable future.
HYPE DOESN'T HELP
Why does it matter? Because hype is costly--both to customers and vendors. Unrealistic expectations hurt everybody; and, in today's increasingly brutal business environment, where speed and scale are paramount, betting on the wrong horse in IT can, if not cost the farm, at least cause it to be put up for auction. Like healthcare, the ultimate responsibility falls on the customer. While we would all like to think that there is a Marcus Welby out there who is more concerned about our well being than we are ourselves, the truth of the matter is that the health of the individual is ultimately dependent upon his/her own commitment and the ability to discern who is the quack and who the talented practitioner.
So, caveat emptor. Complex problems usually have nontrivial solutions and, instead of defining your reality, those visions of integration dancing in your head could easily turn into nightmares if cultivated solely on hype.
Hammer is CEO and president of Evolutionary Technologies International, Austin, Texas.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Industry Trend or Event|
|Comment:||The claims of pundits about how a technology will dramatically change the world have a history of not coming true and often reflect wishful thinking.|
|Date:||Jan 1, 2000|
|Previous Article:||An evolution revolution.|
|Next Article:||Cache and carry.|