Gasoline prices are going way up. Inflation is slowly picking up. The cost of a barrel of oil is setting records. These external events will have an effect on your companys technology spending, but it is difficult to figure out what that effect will be.
Should you dust off plans to increase the number of employees working from home to save commuting dollars? Should you re-examine those technology projects aimed at making sure you use the trucks in your fleet most efficiently? Should you make your technology purchases now while leasing and loan rates are still low?
Despite a long history of technology purchasing by corporations, I find there is still a lack of sound advice on how to get the most mileage from your technology dollar. Technology purchasing by crisis—for example, when the latest computer virus is making its rounds—is not very effective. Technology purchasing as a reaction to large economic trends also doesnt strike me as a way to get ahead of the curve.
Getting the most out of your technology budget was a theme of the MIT CIO Summit, hosted by the Massachusetts Institute of Technologys Center for eBusiness. The daylong forum was packed with CIOs eager to learn about "Enabling the Agile Enterprise," as the program was named. "CIOs face tremendous challenges in todays business environment," the program brochure said in its introduction. "Change is ubiquitous—technology, business models, mergers and acquisitions, outsourcing, compliance, and security, to name but a few areas," the description continued. I was interested to hear how MITs Center for eBusiness—always strong in research—would advise those CIOs to approach that ubiquitous change.
I wasnt disappointed; Ill try to recap the research presented by Erik Brynjolfsson, director of the center. You can find a lot of detail about his findings at ebusiness.mit.edu/erik, but basically his research found that companies that adopt and successfully implement a robust digital structure are more productive than companies that go with either a half-measure or none at all. Id put his findings in the "IT does matter, if you do it right" category. His research brings some factual basis to the ongoing argument of whether and how you should invest in technology. It also buttresses a couple of points I scribbled down while listening to the presentation. Here they are:
Investing in technology for the sake of technology has always been a bad strategy. This type of investment is fading but has not disappeared. Examples of projects pursuing technology without a clear purpose include companies building out wireless networks for no apparent reason or engaging in ERP projects simply because that is what competitors are doing.
Investing in technology to make a broken process go faster is worse than no investment at all. Too often, when you ask technology or business managers what technologys role is, they quickly answer that the purpose is to support a business process. Whats needed is greater questioning of the business process: Is it worth the investment, or will it become a sinkhole for implementation fees? Brynjolfsson had startling figures on the cost breakdown for big projects such as ERP development. Of the $20.8 million a typical ERP project consumes, only $800,000 goes for hardware costs and $3.2 million goes for software costs, including licenses and fees. The biggest chunks of the overall cost of an ERP project go to consulting and training fees: $9.3 million for a nine-month implementation and $7.5 million for deployment.
Brynjolfsson lists seven attributes of companies that get the most productive use from their technology investment: moving from analog to digital business processes, fostering open information access, establishing a distributed decision-making process, linking incentives to performance, maintaining focus and communicating corporate goals, hiring the best people, and investing in human capital. Maybe you should see how your company compares with this checklist before starting your next IT project.
Editor in Chief Eric Lundquist can be reached at email@example.com.