Bernard Golden, CEO of HyperStratus, wrote an excellent article on his 6 new predictions for IT. For once, I really liked this “prediction” article. It is backed with some good experience and foresight.
He predicts that the cost of IT components (not just hardware, includes software as well) decreases precipitously. Referencing Jevon’s paradox, he also talks about how this results in increase in total IT spend( instead of decrease).
As I was reading, it occurred to me that as the confidence in SaaS applications grow, there will be a significant shift to move more and more enterprise IT applications to SaaS. But in the near future, IT still needs to maintain legacy applications. Further, with ever growing SaaS applications in every possible domain one can think of, the cost effectiveness of maintaining legacy applications will be a big question, resulting in tighter budgets.
While the hybrid cloud (“application stretching”) solutions can help IT leverage cloud economics for legacy apps, but they cannot be a permanent solution. In the absence of an equivalent SaaS application, the only long term solution is to look for a green field approach i.e., build the application using cloud technologies. Unless IT upgrades the current app development skill sets, cloud application development is going to be a tough game to catch up with. PaaS platforms could significantly reduce the steep learning curve. Since not all enterprise applications have the same scalable requirements as that of Web2.0 applications, I wouldn’t be surprised to see newer PaaS platforms hiding all the cloud stuff underneath and provide the familiar enterprise development environment/frameworks. Infact, RedHat’s OpenShift is a good example of steps towards that – they are probably the only known PaaS vendor with J2EE support. Microsoft’s Azure is another PaaS platform to watch out for enterprises who invested in DotNet technology.
Another interesting aspect to look at is that as more and more SaaS services are adopted by enterprises, there is going to be definite need for IT to develop custom applications that integrate services from many SaaS applications. One approach is look at solutions like IBM’s CastIron. Another approach would be an interesting opportunity in the PaaS area i.e., a PaaS platform focused on integration in the cloud.
The "Present" I live in 