A few years ago, I was treated to a demonstration of a Fortune 50 company's dazzling new metrics program. Staffers had created a series of dynamic scorecards, which were all posted on the corporation's internal Web site. With the click of a mouse, anyone could review current measures for every group in the business unit.

As the senior executive leading the demonstration pulled up a scorecard on her computer, I pointed to a particular measure and asked, "Is that a good thing or a bad thing?" She looked me straight in the eye and admitted, "Frankly, I have no idea."

That's not at all unusual. Plenty of companies have collected untold gigabytes worth of performance data without giving much thought as to how they'll use the information to manage their business. And many collect it without giving much thought to whether they'd be willing to use it at all.

Before establishing measures for your company, you must answer two key questions:

- Will I change my behavior, or ask employees to change theirs, based on this measure?

- Does the potential benefit gained from this information exceed the cost of obtaining it?

If the answer is no, don't bother. You've undoubtedly heard a million times that "You can't manage what you don't measure." That may be true, but it shouldn't be taken as encouragement to leave no process unmeasured. Though I would certainly agree that you can't manage what you don't measure, I'd hasten to add: "But don't measure what you won't change."

Five easy pieces
To get the most from your metrics, you have to take the measurements out of the realm of the abstract and translate them into something that's meaningful to the people on the shop floor. You must demonstrate to your employees exactly how their performance affects every aspect of your business, and then work with them to select and implement the measures. It's not easy, but it's definitely doable. A number of companies have achieved excellent results by following these simple rules:

1. Articulate the objective(s) clearly. Do your people understand your strategy? Do they understand their role in helping you achieve that strategy?
2. Set specific targets. Let's say your objective is to increase customer service, and you decide that a key measure is the average wait time in the call center. Setting a target of reducing wait time isn't enough. You have to come up with something more specific, like a goal of answering 80 percent of all customer calls in 20 seconds. With a clear target, employees readily know how far they've come, how far they still have to go, and when they've reached the goal.
3. Measure progress against the goal. Some companies get excellent results using Statistical Process Control (SPC) and Six Sigma methods to assess their progress. But other companies have found these structured approaches to be a poor fit with their more "unstructured" cultures. For them, a better choice might be to use informal methods like Excel spreadsheets or even graph paper to plot their progress. It doesn't have to be fancy; it just has to be clear.
4. Don't stop until you uncover a problem's root cause. So far, so good. But now comes the difficult part.What do you do if you're not meeting your goals? We recommend continuous improvement methodologies, such as the Deming Wheel PDCA (Plan/Do/Check/Act) or the Six Sigma DMAIC (Design/Measure/Analyze/Improve/Control), as good guidelines to follow.
   
   For companies that aren't familiar with Six Sigma, a simpler alternative may be to develop a Pareto chart. The Pareto chart is designed to determine the root cause of a problem that is, to find out exactly what's keeping employees from meeting their goal. Assign the workers who actually do the job to carry out the investigation and create the chart. After all, they're in the best position to determine what barriers stand between them and success. Asking "why" two to five times usually yields the root cause.
5. Fix the problem. Then develop a corrective action plan. This is where results begin. Employees who understand what's keeping them from meeting their objective can develop an action plan to fix it.
   
   But the work doesn't stop there. Even the best action plan will accomplish absolutely nothing if it ends up in a stack of papers on a manager's desk. The key to any successful performance management program is to take that information and apply it, to use it to make improvements.

The next time a "metrics report" crosses your desk, take the Metrics Challenge. Ask yourself if your company is a metrics collector, a metrics poster or a genuine metrics user. If you decide you're not a user—that you're not actively using your metrics to drive improvements—at least you'll have the answer to one question. You'll know why your metrics program isn't getting the results you expected.