Return on Infrastructure Employed

Rather than simply managing IT project by project, one company is taking the broader approach of comparing overall technology expenses with revenue...
Russ BanhamMarch 23, 2005

When Arinc Inc. makes decisions about technology projects going forward, it begins by looking backward.

The Annapolis, Maryland-based company, which provides systems engineering solutions to the airline industry, has taken to examining its ROIE (return on infrastructure employed). ROIE is a retrospective comparison of net earnings (for Arinc, about $700 million in 2004) with yearly IT operating expenses — the networks, systems, and applications that underpin the business.

“ROIE helps me examine IT expenses in the context of our overall management infrastructure to see if they lived up to their promise,” explains chief financial officer Richard Jones. If IT delivers on that promise — the case at Arinc, according to the ROIE metric — then Jones is more likely to bankroll future technology projects that have a strong business case. “This is a much more effective metric than simply ascribing an ROI to each IT project and then justifying the individual rate of return,” he asserts. “In my experience, when you try to justify ROI you end up rationalizing a whole bunch of spending without being able to point to a specific result. We have found that an aggregate metric like ROIE is a far more useful indicator of overall IT effectiveness.”

4 Powerful Communication Strategies for Your Next Board Meeting

4 Powerful Communication Strategies for Your Next Board Meeting

This whitepaper outlines four powerful strategies to amplify board meeting conversations during a time of economic volatility. 

Is ROIE this year’s TCO (total cost of ownership), EVA (economic value added), CBA (cost-benefit analysis), NPV (net present value) or what-have-you? Without question, this new twist on an old theme underscores the frustrations of finance executives who must gauge tangible returns from costly technology investments. “If IT costs are going up and are not scaling with our business growth, something’s out of whack,” says Jones. “I wanted some way to continually question our processes and our infrastructure architecture.”

Lies, Damn Lies, and ROI

Technology research firms, consultancies, and vendors have built a cottage industry on helping companies quantify the hard dollar savings and softer productivity benefits of IT projects — and each advisor has its own unique series of measures to justify IT expenses. Gartner, for example, will append a long list of “risk-adjusted” metrics to its ROI calculation; Forrester Research will assess the TEI (total economic impact), including the future operational benefits after the technology is implemented. (See “Two More Twists on ROI” at the end of this article.)

Why tinker with something as solid and simple as good old ROI (net cash flow divided by initial investment)? Because ROI is anything but solid or simple. “It’s as elusive as the holy grail,” says Chip Gliedman, vice president of Cambridge, Massachusetts-based Forrester. “There’s no GAAP-accepted definition, no ROI tool that will tell them exactly what to do, no exact ROI model. As [Benjamin] Disraeli said, it’s nothing but ‘lies, damn lies, and statistics.’ “

The 19th-century British prime minister was surely discussing another subject on the floor of Parliament, but most CFOs would agree that statistics are quite flexible, for good or ill. “ROI is an inexact science simply because you’re calculating the future impact of an infrastructure decision, and the future is impossible to predict accurately,” says Jack Heine, a vice president and research director at Stamford, Connecticut-based Gartner. “You can make assumptions on economic cycles and business growth, which certainly affect the return, but you can’t be sure they’ll prove true. Not that many companies actually evaluate ROI down the road anyway.” Notes Heine, “If you have a project with a five-year ROI, chances are the people in IT that got it going will be elsewhere when the time comes.”

So why bother calculating ROI at all if the outcome is unpredictable? Because “there is a history of IT projects not living up to billing, coming in late, and costing more than they were supposed to,” says Randy Perry, vice president of business value strategy at Framingham, Massachusetts-based IDC. “Also, there are few ways other than ROI to measure the cost of a project versus the efficiency it is expected to offer.”

“ROI is a fact of business,” affirms David Hebert, IT practice leader at the Atlanta-based Hackett Group. “No CFO will make a large investment without having a fully detailed ROI explicating the benefits in financial terms.” Gliedman agrees: “ROI helps you better understand all the issues and circumstances so they can be weighted in light of external factors.” Think about what happens when you visit a discount retailer to “stock up and save” on household goods, he observes. You do a quick cost-benefit analysis, but that may not take into account your short-term financial situation, or how much storage space you have at home, or whether a short-term CD might provide a better return, or the logistics of getting your purchases home. Concludes Gliedman, “You need these other factors to make an informed decision.”

Putting a Business Hat on IT Managers

“The issue isn’t ROI as such, but defining metrics that are useful and meaningful,” observes Michael Smith, a Gartner vice president in applied research. “Everyone wants a tidy, safe and comfortable ROI metric. But it’s the metrics within the metrics that really tell the tale.”

Hence Arinc’s decision to develop ROIE and weigh it against traditional ROI calculations for infrastructure investments. Rather than simply manage IT project by project, Arinc takes the broader approach of comparing yearly technology operating expenses with net earnings. If that first number rises out of proportion to the second, IT may be consuming more dollars than justified by Arinc’s business growth. “ROIE is really geared to the net effect of a collective set of projects,” says Michael McShea, Arinc senior director of global project management, “to see if they are tied to our business objectives.”

“Like all companies, we want our cost growth to be lower than our revenue growth,” McShea explains. Unless that’s true, he continues, “you’re not improving the performance of the IT organization from a financial standpoint. In that case, the CIO is not doing his or her job. You should only undertake projects that improve your ROIE.”

Although ROIE is a retrospective calculation, it helps in deciding if the costs and benefits of a prospective project make financial sense. For example, say the CFO is presented with the business case for a $10 million customer-relationship-management system. By examining the company’s ROIE, the finance chief gains a better understanding of aggregate IT costs and performance during past projects. If the ROIE is less than stellar, it forces the conversation to a solution that offers the functionality of the CRM system at lower introductory cost. “If revenue is growing and EBIT (earnings before interest and taxes) is steady, but the increasing costs of IT are disproportionate to what the business can accept, the company may want to look for other ways for the project to fly, where costs are eliminated — such as a hosted solution,” says McShea. “It forces IT managers to put on a business hat to minimize fixed costs.”

At Arinc, as it turned out, the converse was true: The company’s analysis of its ROIE gave Arinc the assurance to make significant investments in technology even as its main market — aviation — was suffering. “The airline industry is beset by bankruptcies; their problems obviously put pressure on our margins,” says McShea. “Nevertheless, our ROIE indicated we could risk investments in a new billing system and network that we would have thought impossible in the kind of market we’re in, had we not developed this metric.”

“We’ve spent some money and are now more productive than we used to be,” Jones chimes in. “But we wouldn’t have spent the money if we felt that IT was not in synch with our business objectives. The ROIE validated the investments made [by IT].”

Arinc’s ROIE indicates a 10 percent improvement in IT’s return on dollars spent in 2004, compared with the previous year. “At the same time, IT spent 15 percent more,” Jones says. “The moral is that returns are not always about cost reduction.”

Two More Twists on ROI

Where ROIE is retrospective, TEI is prospective, and it, too, can provide another piece of the technology-decision puzzle. Forrester Research offers to analyze the “total economic impact” of a contemplated project in terms of other projects that may spring from it, integrate with it, or otherwise benefit from it.

“If a company is planning to implement a new application server infrastructure to support a specific project, it would make sense to ponder the possibility of other similar projects that come down the road that have significant value,” explains Forrester research analyst Andrew Bartels. Quantifying those other projects, he continues, “allows you to assert a probability of 20 percent or 50 percent lower upfront costs” on the project that’s already under way. In effect, maintains Bartels, “you’re buying an option — and the option can expire on exercise and out of the money — but it is there as a hedge to give you more flexibility.”

Bartels’s colleague Chip Gliedman elaborates: “With a financial option, one purchases the right to acquire a stock at a price negotiated today; in the same regard, investing in additional infrastructure above today’s needs can enable the deployment of future applications. In many cases, these applications may not yet be identified or budgeted, but the right to take these actions in the future still has value to the organization that, typically, is not reflected in traditional ROI efforts.”

Stanford University has used TEI in its calculations for every IT project undertaken since 2003. “We just used it as part of a project we implemented for doing patches on our Windows machines,” says Bill Clebsch, Stanford executive director for IT services. “We realized a number of future options that were a byproduct of the security software, such as the ability of department administrators to ‘push’ out new desktop images or software when their department needed it, rather than have them come to IT to ‘pull’ it. The value of that future option was almost as large as the security gain.” In addition to a 20 percent ROI and expected savings of $700,000 over the next five years, Clebsch adds that Stanford could potentially save another $1.2 million via the future option.

Gartner offers a risk-adjusted take on returns from technology that looks at “the flip side,” explains research director Jack Heine: the prospect of “the IT project not achieving the ROI.” To do that, Gartner assembles a business performance framework of 54 metrics affected by the project, calculating such factors as the effectiveness of suppliers and the sales department, and the responsiveness of the finance department, the markets, customers, and regulators. Within each category, Gartner drills down to illuminate potential risks: In evaluating customer responsiveness, for example, the consultancy quantifies on-time delivery statistics, order fill rate, service accuracy, and material quality, among other factors.

These metrics are like an “EKG on the enterprise,” says Heine’s colleague Michael Smith. They “focus the attention of both business and IT professionals on issues that generate future cash flows: things like demand management, supply chain management, and support services.”

Armed with this data, he argues, a company can “capture the total effect of an investment in IT. You can now talk about the effect of a new CRM system on on-time delivery because you now have a metric that allows the discussion…you can describe the system in the context of improving the metric.”

“What keeps CFOs up at night,” observes Smith, “is a $10 million IT project offering many benefits, all of them intangible. We’re trying to give as much substance as possible to the insubstantial.”