For many companies, the IT group is a cost center to be minimized. Ironically, the cost of IT systems remains flat (actually falling on the basis of cost per unit of capacity delivered). It’s the cost of administration, management, operations, and power that keeps going up.
Steve Mills, Senior Vice President and Group Executive, IBM Software & Systems, made this point at a recent briefing: “Server management and administrative costs keep going up and the cost of servers drop and the cost of power keeps rising. These are the (IT) economics facing business today.”
The economics Mills cites are changing the way IT systems and capabilities are bought and deployed. The challenge to rein in IT costs has, in part, fueled the interest in cloud computing. Many organizations are hoping the cloud will bring IT spending under control. Forbes Magazine looked at the issue here and found that there isn’t a simple answer.
The problem as Mills lays it out is that of the three main buckets of IT costs—hardware, people, and power—only one is flat or dropping. The other two continue to go up.
Cloud computing can change the economics to some extent, but it won’t eliminate the overall corporate IT spend or even much reduce it. What it mainly does is shift the IT spend from CAPEX, by eliminating the need to invest in new systems, to OPEX by paying for IT capabilities and capacity as services when needed.
Cloud computing certainly can be useful to augment corporate IT. The public cloud saves a company from investing in IT capabilities that may be used only occasionally, such as to meet a spike in demand. It also is demonstrating real value for backup and disaster recovery. In that sense it changes the economics of IT.
But as Mills noted, the big economic burden does not come from a handful of servers sitting around idle while awaiting the next spike or older systems relegated to backup and recovery roles. The biggest economic toll comes from the cost of people.
Traditionally, IT groups were staffed to acquire hardware and software, connect and integrate it, and support it going forward, changing things as needed when the business changed while supporting users along the way. To do this, companies had to hire and retain people skilled in hardware platforms, applications, system integration, data management, and more. Even with the growing interest in cloud computing, the traditional approach remains the dominant model despite getting more expensive each year.
One proven solution is automation. Systems automation allows the IT group to support more servers, more storage capacity, and more networking with fewer administrators. Policy-based automation combined with the cloud can enable user self-provisioning of systems, enabling systems to be provisioned and deployed in minutes. A small core of admins handle the exceptions. BottomlineIT will take up automation in an upcoming piece.
Another possible solution, Mills suggests, is adoption of a factory model. In the factory model, the company doesn’t buy the system piece parts and cobble them together but buys ready-to-use systems pre-assembled, pre-integrated, and pre-optimized by the vendor. The vendor has the tools, experts, and process efficiencies to do this better and cheaper. Like automation, the factory model can impact the IT people cost bucket.
IBM has been offering such factory built integrated and optimized systems as early as 2008 for integrated SAP hardware/software systems. Lately, it offers the Migration Factory for companies that want to move to new, more efficient IBM systems and its first factory-built system packaged as an appliance, Netezza, for data warehousing.
Other vendors offer versions of the factory approach. Check out Oracle’s migration factory. HP offers the HP server blade system factory. Dell offers Custom Factory Integration for hardware, images, applications, peripherals, and documents.
The factory approach, especially when combined with automation and the cloud, starts to change the economics of IT by reducing the amount of pricy IT expertise a company has to hire and retain. The tradeoff is fewer IT options offered. Still, by selectively combining the factory model with automation and the cloud, the CIO finally can begin to impact the economics of IT.