My client D.K. asks: “What terminology can we use to understand the different types of performance measures and KPIs?”
We need a consistent terminology for our Performance Frameworks to be strong enough to support a large number of measures. But it can easily get messy.
We refer to performance measures using all kinds of names: KPI, PI, lead indicator, lag indicator, metric, index, key figure, to name a few.
And we classify performance measures using all kinds of buckets: financial versus non-financial, qualitative versus quantitative, strategic versus operational, to name a few.
As I said, messy. What we really want is a simple and consistent method of naming and classifying our measures. That way, it’s easier to understand how any specific measure fits into the overall Performance Framework. And it helps us get balance and alignment in the overall Performance Framework.
Please don’t assume that I have the complete answer here. But what I will share with you now, I hope will give you some sample dimensions for how to create a performance measurement terminology that supports your organisation’s Performance Framework:
Financial measures are any measures based on quantifying dollars. The most typical examples are Profit, Revenue, Costs, Return on Investment, Equity Ratio and Cash Flow.
Non-financial measures are based on anything other than dollars: attitudes, time, weight, widgets, events, and so on. Customer Satisfaction Rating, Delivery Cycle Time, Landfill Mass, Camaro Production Rate, Workplace Accidents are examples.
I’m not sure if financial versus non-financial is a useful distinction to make. Dollars are just one possible unit of measurement.
TIP: In your Performance Framework, make it clear what the measurement units of each measure are.
Quantitative measures are based on numbers, where those numbers are gauging the size or amount of the result you’re measuring. These numbers can be discrete counts or continuous values that include decimals.
Truly qualitative performance measures don’t exist. Qualitative data is based on non-numeric data, and that’s usually in the form of words. Gender, geographical region and mode of transport are examples of qualitative data. They help us analyse patterns and causes in our measures, but they aren’t themselves performance measures.
TIP: Distinguish your measures as either continuous or discrete, rather than qualitative or quantitative.
KPIs, key performance indicators, key figures… What exactly does ‘key’ mean? And if it’s not ‘key’ then what is it? Likely ‘key’ means important. So your key measures or KPIs are those measures that are most important to monitor and improve right now, given your strategic priorities.
But ‘key’ might also be a label that distinguishes your business-as-usual measures from those measures that are tightly linked to your strategic goals. You want to improve your key measures, but just monitor and maintain your business-as-usual measures.
TIP: Define what you mean by ‘key’ if you’re using it to name any of the types of measures in your Performance Framework.
Lead indicators are performance measures that monitor results which have some predictive power over lag indicators. A change in the number of near miss accidents in the workplace might be a lead indicator for the lag indicator of lost-time injury rate.
To be a real lead indicator, a measure must give you clues about what a lag indicator is going to do BEFORE it does it. The idea is to use lead indicators to prevent problems, not diagnose them.
You need to do some decent cause analysis to find the best lead indicators. Some might be obvious, like the example I gave above, but others are more subtle, and only revealed when you dig into the patterns and correlations in your data.
TIP: Add relationships between measures to your Performance Framework, to document the important lead-lag and cause-effect relationships.
This is a distinction that is contentious. More often than not, the terms measure and indicator and metric are used interchangeably. And that’s how I use them (because I’d rather focus on good design and use of measures, and not a universally accepted terminology). But here is some food for thought if you want to be a bit fancier than that:
A measure is a direct gauge of the result you’re monitoring, just like Average Delivery Cycle Time is actually based on data directly about actual cycle time.
An indicator is an indirect gauge of the result you’re monitoring, just like Average Likelihood to Recommend is based on intentions of customers to recommend your service, not their actual behaviour of recommending.
A metric is defined by my friend Marty Klubeck, in his book "Metrics: How to Improve Key Business Results", as "a picture made up of information, measures, and data… Metrics can also include other metrics… a metric tells a complete story, fully answering a root question."
TIP: If you’re going to use different words to name types of performance measures, write a glossary to define the difference.
An important part of any Performance Framework is some kind of heirarchy among performance measures. Not a heirarchy of importance, but rather a heirarchy that helps anyone throughout the organisation understand how the measures they affect ultimately affect the success of the organisation.
The primary linking mechanism is such a heirachy is cause-effect. A good way to build this is to start with the performance measures that monitor the ultimate success results of the organisation: the vision, mission and current strategic goals.
Then, examine how each of the core business processes impact these measures. The process results that have the impact will then be measured.
Next, examine how each activity or sub-process within the core processes impacts on the process results. The results of these activities or sub-processes will then be measured.
The resulting set of measures will form a heirarchy. The number of layers in this heirarchy is somewhat dependent on the complexity of your organisation, but 3 to 5 layers is usually plenty.
TIP: Design the heirarchy of your Performance Framework, and make sure each measure is located at the appropriate level.
What other ways do you classify and name your performance measures? Let’s collaborate to identify as many options as we can, so we have more choice when we design our KPI terminology. Share your suggestions on the blog.
I like the idea of a glossary. I believe there are a lot of people throwing out measures, metrics, standards, etc. that don’t understand the nuances involved. A glossary pinpoints exactly what you are reporting and takes the guess work out.
I’m in the process of creating a “KPI Library” with KPI name, description, calculation and which way is positive (increase is good or decrease is good)
Hi Patrick, i am interested in how you are getting on with your KPI Library etc.
Hi, can anybody share their own KPI library they created. Im keen to see how you link from top level strategic KPIs down to operational. THANKS
please can someone send the me kpi library to firstname.lastname@example.org
I really need them.