Measuring is important in process improvement. You've probably heard that. Maybe a few times. The problem I have found is that the people who are the ones saying it may seem to believe it, but they don't seem sure why. That often leads to a common misdirection with process improvement programs: thinking you have to support it with a complex measurement program.
When people are led down that path, the result can indeed be a very complex measurement program, maybe even a good one, one that on paper promises to reveal a host of nuances and facets of your process set. But then comes the chore of doing the measuring: finding someone to amass and collect the data, analyzing and figuring out what the results might indicate, reporting and communicating the results, and then deciding what to do with it all.
What happens next—after the paper program is admired—is often nothing. It takes work to collect measures. It takes energy. And resources, and time, and (admit it) interest. When an organization drifts from the chief purpose and use of process program measurements and creates a mathematical and statistical Methuselah, no one wants to be bothered with it. No one wants to look at it. It appears too cumbersome, too tangled, or like too much work. Besides, people might think, why are we collecting all this data anyway?
Measuring is important in process improvement. But the reason is simple. Measurements tell you how you are progressing. They help show you the direction you are moving in. A trip odometer is a measurement system. A compass is a measurement system. Kricket Ichwantoro and Nicole Whitelaw, the process quality analysts I try to work with whenever I can, know the measurement conundrum well. They often step in to help IT shops plan for and develop metrics for new or existing process programs. Their common philosophy is pretty straightforward: keep it straightforward.
The measurements you define will work best when they possess three traits:
The measures tie to hard business objectives.
The measures give you meaningful information.
The measures are things your people can identify and collect.
With those traits in place, you don't need a complex measurement approach to help analyze your process program. If the approach you design reflects real business needs, if it contains data that means something to the organization, and if it focuses on data that's readily available, you can go a long way toward setting up the foundation of what can become, with use and improvement, a solid metrics-analysis program.
To begin with, though, two things are handy to consider: measure to monitor and measure to know.
The concept of measure to monitor sets the stage for the opening advantage metrics can deliver. Regularly measuring your process program is an effective way to monitor the program's affinity with business objectives and project efficiencies.
As mentioned in Chapter 3, one of the key factors in establishing an effective process program is to tie the activities in your program to overall business objectives. It's important to shape your program to the needs of the business, and it's important to demonstrate that your program is helping your organization achieve those objectives. And so, as you decide what kind of measures you might begin to trace through your organization, think about what you'd like to demonstrate.
Is fidelity to schedules a point of special interest? Is the rate of resource churn of particular concern? Budget? Length of project phases? The number of change requests?
All of these kinds of metrics—quantitative snapshots of evolving activity—can be used to help you monitor how your projects are doing. Not just in retrospect, but in real time as your project unfolds, so that you can better manage the unfolding.
Measuring can also help you monitor the performance potential of your processes. Not just how things are working now, but how things will probably work later on: process performance. Even if your measures begin simply in a qualitative mode, you can begin to discern how well the processes, procedures, and other tools that you've implemented are working for your teams.
With time, this facet can become the greatest growth factor of a process program. The ultimate achievement is to measure the performance of your processes to such a degree that your people, by implementing the processes, can predict quantitatively how they will perform and thus anticipate how the overall project will perform.
That's quite an achievement. And many process-centric companies have attained that achievement. But for now, for what we are after—sustained process improvement—a little less is just as respected. You needn't begin with what you know you know. It can be just as valuable to begin with what you'd like to know.
Over time, as your measurement repository grows, the information that begins to accrue will point toward opportunities for improvement. The repository will give you a foundation to know what to improve. The following is an example.
If you're monitoring how well your teams comply with published processes, and your auditors are recording noncompliance issues, their measures might indicate that for one particular process, noncompliance is abnormally high. This might indicate a few things. Maybe your training needs to be beefed up for that procedure. Maybe team members haven't been trained to effectively follow the process. Then again, maybe the teams are regularly avoiding the procedure because it doesn't fit the way work flows. Or maybe the way it's built does not allow for the kinds of tailoring that's needed for the particular activity across different projects.
So measurements, carefully designed to reflect your business and your program, can help you know what you might need to improve.
Measurements can also establish avenues to improve elements of your program in specific ways. As you begin to look at ISO 9001, CMMI, and Six Sigma in Part 2 of this book, you'll see how these models place particular emphasis on metrics. And with greater levels of capability and sophistication come increased measurement. The idea these models support is that metrics provide valuable information for making data-driven decisions. In other words, your experience, judgment, and intuition are valuable assets for your process improvement strategies. But a strong complement to these is the solid component that data supplies.
Measuring the right things in the right amounts will help you sustain process improvement activities by providing you with hard data that can complement your experience and judgment when it comes to refining the program for more efficient use in your organization.