With investment banks reporting layoffs and tightening their belts in light of the economic downturn, one of the major themes emerging in IT circles is sustainable cost reduction. “The industry does carry a lot of cost,” says Bob Gach, global managing director capital markets at Accenture in New York. “The industry hasn’t had the courage or need to attack these costs,” says Gach.
When the industry was generating huge profits from packaging and trading complex derivatives and needed to launch a system in 30 days, cost didn’t matter. But now that the bubble has burst, the industry is more focused on not only how to reduce costs but sustaining the operating model in good times and bad.
The concept of “sustainable cost reduction” takes a more holistic or systemic approach to cost reduction, according to Accenture’s spokesman. Instead of one department looking to reduce costs and another department working independently, all the departments work together as an enterprise to reduce costs. This makes the cost reduction more sustainable in the long run, says Accenture’s spokesman.
Check out this interesting post which is talking about IT investment within the finance sector, it’s an interesting read and does make some good points. I wonder if we don’t need to turn the way we do business around? By that I mean consider smaller leaner more expensive teams that deliver, a team of core individuals that might only perform one task, but do that task to a high standard? In the virtual or by that matter the physical server world, how many engineers should you have to run your server estate? On the basis that it’s 300-500 Windows servers per engineer, would it not be cheaper and more effective to have 5 engineers running that 2000 server estate? You might have a separate server build team, or patching team, but 5 key guys that know the server estate, that are accountable to production issues - that maintain the dialog with the application support teams, with the business sponsors?
How do we want to do IT? Do we want to lower costs long term? Is the aim to shift the capital cost to the operation costs - by reducing the capex costs does it not indirectly increase your operational costs - older servers cost more to run, typically run older operating systems which require more hands on support/maintenance. Stepping back for a second, previously we were averse to spending money, to capital cost, we’re now moving to operational costs, reducing the cost of support, the costs of deployment etc, at what point do we say stop? Where do we draw the line and state, the cost for 24/7 operations is X, for week days, 9-5 it’s Y, you choose. Are we going to move to a tier’d infrastructure, an infrastructure or data center that is designed, deployed and built for it’s target use, it’s target market - availability needs to be 80% - we’ll run the servers hotter and deploy less cooling?
