Hardware will be produced by a small number of vendors that will fight over increasingly shrinking profit margins. Services will be provided by the same silver-tongued devils that provide them now. Software will confuse the hell out of everyone.
In the old days we wrote code. Then we installed it. Then we rented it. Eventually well assemble it. But code still doesnt like other code, in spite of all of the progress weve made with interoperability and integration. While weve made progress with standards, theres still a lot left to do.
The Golden Rule Of Business Technology Investments
How To Avoid An IT Project 'Death March'
IT In 2007: Budget and Trends
Gartner to Discuss SaaS Explosion at Show|
But lets assume that well get it right, that software will work together reasonably well. How will we acquire and deploy it?
Over the past few months Ive asked a number of CIOs and CTOs a question: if they had a technology do-over, would they still install their enterprise application?
Not one of them said they would. Why not? Because it took them all years to get the software to work and in some cases the projects cost hundreds of millions of dollars. Some of these CIOs got fired when they exceeded budgets and schedules; others struggled to realize the benefits everyone promised when they signed the contracts.
Some of the same CIOs and CTOs told me that they werent interested in open source software because it was too flaky and they didnt want to be associated with the open source crowd. What?
Some of them dont trust Web Services and think that service oriented architectures (SOAs) are still ideas, not reliable software architectures.
Some dont trust Mark Benioff or IBMs on-demand guys.
All of that said, heres how it will go:
Big time CIOs will not launch multi-year, multi-million dollar software implementations. Theres too much time, money and politics involved, and many of the biggest projects havent delivered the goods. Only CIOs in the last year of their employment contracts will attempt multi-year software projects.
If at all possible, CIOs will rent versus buy-and-install major software applications. Theyre all secretly hoping that ASP 2.0 is successful. They really dont want to get back into the enterprise software acquisition, deployment or support business. Most of them are really bad at it and they just dont have the stomach for technology marathons anymore.
CIOs will pilot as many SOA implementations as possible to determine where the price/pain/performance ratios lie. They really want all this stuff to work they need software-as-a- service (SaaS) to become a reality (but theyll be happy to sit on hosted applications for as long as they can as long as someone else is hosting them).
Software-as-a-service will take much longer to evolve than anyone thinks, but eventually will it will mature to the ultimate mix-and-match software architecture.