A Failure of Cost-Benefit Analysis


By John Parkinson

I am always looking for ways to make people more productive. Often, that means eliminating wasted time from their working days. I remember one example for quite a long time ago, when I found a situation in which team members were scattered across many separate locations over several floors in a building.

By moving everyone around to a set of common locations (took a week to plan and a weekend to actually move) we saved almost 10 percent of their working day that had been spent moving around to meetings or to find people who didn't answer their phones because they too were walking around looking for people who...

There is good research to support these kinds of improvements, including research on the benefits of large display monitors, higher network bandwidth and faster hard disk drives. Sure, these are all more expensive than their smaller and slower equivalents, but the time gains more than offset the additional costs over a very short period. In some cases you can give people back as much as an hour a day in productive time because they wait less for a PC to boot or spend less time navigating around a spreadsheet. My own data suggest that the savings are around 40 hours a year for an average information worker--and that's net savings. You get some breakage because people don't work flat out all the time and actually need some slack in their day.

Still, 40 hours a year translates into a couple of thousand dollars in improved productivity for an average employee who uses a PC most of the time. Across dozens or hundreds or thousands of workers, that should be a simple decision to give them better equipment.

Except it almost never is.

Go to most offices and you see the majority of people squinting at 17-inch monitors and with plenty of time for a coffee run while their PC boots up. Somehow the "cost" part of CBA (which is real and immediate) overweighs the "benefit" part (which is softer - although still measurable - and spread over time).

It doesn't matter that the extra $200 for a 24-inch widescreen monitor will be repaid in less than three months. Or that the additional $300 for an SSD in a laptop will save many people's time in every meeting where it is used as a projection source. Somehow the accountants don't seem to get their own math.

Oh well, we've been here before. Remember when no one really needed a color monitor?

John Parkinson is CIO of TransUnion. To read his columns for CIO Insight, click here.


2 Comments for "A Failure of Cost-Benefit Analysis"

  • Brian Reed December 07, 2009 12:37 pm

    Indeed 'hard costs' vs 'soft costs' and 'gut instinct' vs. 'actual proof' are challenges of cost-benefit analysis and ROI attempts. Over the years I've found that most companies don't have a handle on the true fully-loaded baseline costs, making the case even harder. So we work with clients to help them uncover their own cost baselines, then make the case for benchmarking vs. peer results on where and how the cost savings have been realized. We have found that an educated IT team with their own internal numbers (once they know where to look and typical ways to calculate) can make a more compelling argument, backed by peer benchmarks. but even more important is to go measure afterward and adjust your actual cost baseline. Now you will have a more accurate baseline for future purchases and you will have built credibility with management and finance, which will make it easier on future asks.

  • Cliff Berg November 25, 2009 6:23 pm

    One problem is that this kind of cost-benefit analysis is not convincing. For example, assuming that a company does not spring for a faster-booting PC, who is to say that people do not make up for the small amount of extra time by working a few minutes longer each day? Or assuming that faster-booting PCs are purchased, that people will not squander the time saved by surfing the Web a little more? And I for one don't shut down my PC each day: I put it in standby mode, so the argument doesn't even apply. All these uncertainties put the purported ROI into question. To make a convincing ROI argument, one must have tangible evidence that the ROI will actually be realized. Otherwise, execs will rely on their intuition, and the fact that "soft" ROI claims are often not believed probably indicates that intuition says that they are not to be trusted -- probably for good reason. - Cliff

Leave a Comment