A Lesson in E-Biz Survival: Measure or Else

You might think it would be easy to convince IT managers that it's important to measure their group's effectiveness supporting e-business

You might think it would be easy to convince IT managers that its important to measure their groups effectiveness supporting e-business. After all, as e-business has exploded, consuming a bigger and bigger chunk of IT resources, gauging exactly how efficiently youre deploying those resources would seem like good business.

But youd be surprised. For the past several years, we have been involved in the practice of measuring IT performance by applying many different types of metrics. Convincing our clients that measurement is a key aspect of success has been an uphill battle since we started.

One particularly memorable experience that illustrates the advantages of measurement as well as the challenges of gaining buy-in happened late last year at a company dependent on e-business for its livelihood. Two different project managers at this company were in charge of different e-business development activities.

The first project manager, whom we shall call X, understood the value of measuring all aspects of the development process. She carefully tracked time, resources, defects and so forth during the development and testing of new e-business applications.

After launch, X continued measuring by watching user activity and using this information to improve the site and help marketing people better focus their efforts.

From project to project, she was able to track performance and manage her team with real quantitative evidence of how well it was doing and how effectively the site was performing.

The second project manager, whom we shall call Y, didnt bother with metrics. She was convinced that resources allocated for metric collection would be better utilized in development and testing.

She believed her team was doing its job because it was generally completing projects on time. But, of course, she couldnt prove this.

While Y disdained measuring her own groups performance, she did attempt to gauge user activity and Web site effectiveness. She had a team member run a Web-analysis tool and then provided the raw output to the marketing group.

Of course, that output often requires some serious translation for nontechies, but Y wanted to be off and running on the next project. "They can read the manual and figure it out themselves. ..." Of course, they ignored the data because they could not understand it.

So how did we get through to Y? Well, we would love to take the credit for breaking through to this anti- metric diehard, but heres what actually happened: X began taking on more and more of the large, critical projects, and Y was assigned the mundane, low-exposure projects. Eventually, X became CIO, and Y left the company, providing us with the best example for our slogan: "Measure or else ... ."

We have experienced many successes and some setbacks in our efforts, but fortunately, as time goes on, the historical data that ongoing metric collection provides speaks for itself.