Getting the Most for your Money
By calculating ROI for their new initiatives and then looking back once a project is completed to see if they hit their numbers, Intel's IT department figures it delivered a whopping $1.7 billion (yes, with a 'B') in business value to the company.
Although they won't hit this number in 2006, they also reduced per-employee IT costs by eight percent by keeping costs flat while the number of employees went up, said Martin Menard, director of Intel's High Performance Computing group. Although not officially in his title, Menard is a group CIO who controls $350 million of Intel's $1.1 billion IT budget and is a direct-report to Intel's CIO.
Even more impressive, the business value numbers do not include the return from projects completed prior to 2005 that are now considered just part of keeping the lights on and running the business. The business value calculation is just for new initiatives completed in 2005.
"It's the return on the investment made for new capabilities delivered in the course of a year," said Menard. "So anything to keep the business running (KTBR), we wouldn't track as a business value number."
What does get tracked is project-specific but Menard works closely with his project teams and two types of professionals not generally associated with IT departmentshuman-factors engineers and ethnographers (people who study people)to come up with measurable ROI metrics. They then use these number to go back to check up on their pre-implementation assumptions to see if they were right.
"It's all based on financial modeling for value and return on invested capital," said Menard. "The benefit for a company is it allows you to have a common language to make decisions."
This is basically a form a benchmarking and nothing new, yet Menard is surprised by how few companies (even other groups inside of Intel) don't do this.
"IT shops are challenged just like any business is about the investments they are making," he said. "And, because IT is a total cost of sales (like SG&A), everybody's trying to reduce their total spending and allocation. So this is a tool we've found to be extremely valuable in going back to our customers and describing to them what we actually deliver for them."Menard's efforts are based on the total quality management (TQM) movement that many U.S. companies dismissed and Japan embraced just after World War II (and later used so effectively to dominate the foreign car market in the U.S.). Intel embraced TQM in the '80s to deal with quality issues.
Using TQM in this manner means three things: discipline, data and analysis.
"Good discipline says you collect data, you analyze it, and then you make decisions," said Menard. "Business value it just thatit's project management 101. It gives you a chance to make a better decision verses a gut-reaction. And then you can measure it after the fact to see whether or not you did get the kind of results you expected."
To handle all of this tracking and analysis may seem like a daunting task requiring a huge department but Menard does all of this with just 10 or so full-time staff and the part-time efforts of some 100 project managers. This is a lot compared to just last year when there was only one, full-time person devoted to the value program.
"So it's not that big an investment. I think sometimes people get scared off and think, 'Oh, business value is so hard, we don't have time to do it'; it isn't that heavy lifting."
To get over your reticence, start slow. You can't measure everything, advises Menard, so focus on those factors that will be the easiest to tease out. You will also have to include your finance department to keep track of hard costs and the business units to make sure value is indeed being delivered.
And, until both of these groups sign off on the data, and finance does the math, only then can IT claim the numbers for themselves. "We can't take credit for a business value result until the customer and the finance group say, 'We agree'."