From Measurement to Management
Recently, I've been thinking a lot about process performance management. My experience is that (and I appreciate, dear reader, that this almost certainly doesn't include you) most organizations do too much of it badly rather than much less done well.
Authentic performance management is the secret sauce of effective and sustained process-based management. Notice that I said, "performance management," not just "measurement."
Here are 14 thoughts about how to move from measurement to management.
1. Managing expectations
Measure for the right purpose, a purpose that is widely understood.
Measurement just for the sake of measurement, just to tick the box, is waste. There must be clearly-understood and strongly-argued reasons for a regime of effective process performance measurement across, at least, the high-impact processes. It is not good enough to simply assert that it's "obvious that we should."
Then for each process we are tracking, we must have a clear and shared understanding about which indicators will properly define performance, what a variation from the expectation will look like, and what impact that would have on organizational performance.
We aren't looking to detect perfection; we want a measurement scheme that will find any signal that a change of consequence, or lack of an expected change, has happened or might be about to happen.
2. Scope & focus
It's not possible to actively manage all the processes.
Let's not pretend that we are going to actively manage the performance of thousands of individual processes. That's not going to happen and if we say, or imply, that it will, then the whole performance management idea just lost credibility. It becomes an exercise in creative writing and charting — maybe fun, but hardly management.
What are the high impact processes? Pick five and when you properly understand them, pick another five. Repeat.
3. The critical few
Focus on the small number of things that are contributing the most impact.
Everyone has a version of the phrase "the vital few and the trivial many" given to us by J. M. Juran. Everyone can tell you about the Pareto Principle, or at least the 80/20 rule. We know the aphorisms and yet too often we are focused on the many, not the few; on the 80, not the 20.
In our performance management space, we need a bit more Marie Kondo and a lot less compulsive hoarder. Manage the performance of the minimum number of processes using the minimum number of KPIs that have the minimum number of targets.
4. 100% — Zero
If you set perfection as a target is it really a target or just a noble aspiration?
If 100% delivery or 0 defects is a real target, i.e., your average performance is to be perfection, you need to design and run an error-proof process. Maybe 99.99% is OK because in the real world nobody's perfect, right?
Of course, if that's the success rate in delivering children home from school, the 0.01% error rate might mean that we 'only' lost one child all year — KPI achieved!?
If we set a target that is impossible, or is perceived to be so, we encourage people to do one or more of the following:
a) Ignore the target because it's not real.
b) Give up trying, but keep pretending.
c) Manipulate the data, analysis, reporting to apparently achieve the impossible.
5. Strategic, functional, or process.
To understand process performance, we must be measuring processes.
There are many things we can measure in our organizations. Feel free to measure all of them if you feel the need.
However, if we want to understand process performance, we must be sure we are measuring processes, not functions, people, systems, or some other artifact. What is the process supposed to deliver? If the key stakeholders were happy with its performance, what would the process be delivering and how would we know?
6. Unintended consequences
Knowledge that performance is being monitored may change that performance.
Processes are complex systems working in an ecosystem of complex systems. A small change can have a large impact. We all react to being consciously measured, or even thinking we are being measured. Introducing a process change or even a new process KPI can have unexpected effects.
7. Whose job is it?
The organization chart says nothing about who should manage across the chart.
If you check your organization chart, you'll find that it is nobody's defined job to collect, analyze, report on, and react to measures of cross-functional process performance. Unless there is a practical process governance regime, even if we start well with process performance management, it won't end well.
To give any meaning to process performance management we need to first provide a management add-on to correct the horizontal vision problem.
8. Measures change
Process performance measures will change over time and with usage.
Performance requirements change over time. Processes are changed over time. The operational context changes over time. Opportunities for improvement change over time. Competitive pressures change over time. Society changes over time. You get the picture … why wouldn't measures also need to change? Process KPIs and targets are not constants and should be continually reviewed to ensure ongoing relevance.
9. Continuous measurement
To continuously manage and improve, we must first continuously measure.
Process performance management needs a plan. While it's not something you want to be doing every moment of the day, neither is it an 'annual exercise' designed to see 'where we have been and how we are tracking'. Done properly it gives you insight into where you are probably heading.
Let's work forward supported by the knowledge of what has happened and insights about the predictability of the process.
Measure and analyze in line with the natural cadence of the process. Within reason, taking measures more often, e.g., daily or weekly rather than monthly, will give more insight and better opportunity to intercept the future.
Continuously listening and reacting to the voice of the process will power high-quality decision making.
10. Measuring processes, not people
Process measurement helps improve a process, not find fault with people.
Engage staff to improve processes; don't engage processes to criticize staff.
The problem is almost always with the process or its management, not the people executing it. Setting arbitrary targets, even well-intentioned ones, and either shouting at or praising people will not get performance from a process better than that of which it is naturally capable. If you want to improve the process, change the process.
11. Learning to love measurement
If not love for process measurement, then at least respect for its power.
OK, hoping to convince your colleagues to fall in love with your new process performance measurement and management system might be setting the bar a bit too high. What you can quite reasonably aspire to though is to gain their respect for it by demonstrating how it helps solve, indeed avoid, real business problems and to uncover additional opportunities for improvement.
Effective process performance management is a powerful business tool, not just another burden. To paraphrase Dr. Donald Wheeler, management needs prediction and prediction needs knowledge and knowledge needs signals filtered from noise. That's a process performance management system.
12. Measure outputs, not actions
If actions don't give measurable outputs, we should eliminate them.
Consider the process Produce Report. It might be argued that on-time production of the report is an important process outcome, but successfully producing a useless report on time could not be a good outcome. Also, a focus on timing may result in a reduction in the quality of the contents. If we are to actively measure and manage this process — and we may decide we don't need to — we might find that a more useful metric relates to use and usefulness (which might have an element of on-time availability). Make sure we are still measuring something about the performance of this process, i.e., measure something about the report and its production, not its contents.
There is always a measurable output to measure rather some mechanical action, or if not, stop executing the pointless process.
13. Keep it simple — no fuzz, no fuss
Bring simplicity, not complexity, to management.
There aren't any managers out there wishing "I wish my job were more complicated. I need some new vague and fluffy ideas to make life harder."
Let's keep it as simple as we can. Good process performance management makes the process of management more efficient.
Avoid fuzzy definitions of processes and their KPIs and targets. If there are frequent misunderstandings about the meaning of any definition, then the definition, and/or its communication, needs some work.
Process performance data needs to be reported fully and transparently with plenty of context and meaningful analysis. A Process Behavior Chart will be much more meaningful and useful than a table of percentages or comparisons with last month or last year.
14. Deliver, deliver, deliver
We must be offering a new solution, not a new problem.
As with all of our work in process-based management, the bottom line is that if we aren't delivering proven organizational performance benefits, benefits that are recognized and valued by the organization's managers and decision makers, then we are wasting a lot of time and effort. By our own definitions we are waste to be eliminated.
What problem are we trying to fix? Did we help to fix it? Says who? Prove it!
From measurement to management
If we aren't measuring, we aren't managing, and we can't know if we are improving. Process performance management gives a sound evidence base for decision making. There are many processes and decisions. Without an efficient, effective structure for process performance management we'll be swamped by an avalanche of suspect data and creative excuses.
# # #