Different Types of Metrics

November 2005

We in finance are very, very good at measuring financial results. I regularly teach a class on financial statement analysis where we cover more than 80 metrics in one day. (Sounds like a fun, fun day doesn’t it?)

Finance and accounting weenies can run and calculate financial metrics all day long because we have the data to support the ratios and metrics. We have a well-developed gathering and reporting system in accounting.

When we leave the realm of accounting and finance, the data is a little harder to come by and the metrics are a little more abstract (Yes, I acknowledge that for many of us, financial metrics are abstract…but you haven’t seen anything yet!)

For instance, let’s say we want to know how well our human resource department is performing. Hmm. That is abstract and hard to measure. So for the next few e-zines, I am going to discuss performance measurement—theory and practical application. Maybe something in one of these e-zines will light that yellow bulb in your head and you will experience a Sylvania moment.

Last month we discussed the balanced scorecard model—and we are not through with that, yet. But this month, I want to talk about one of the building blocks of performance measurement—the measures themselves.

Measurement is more nuanced than just counting stuff

When I worked for the State of Texas, the State implemented a budgeting and measurement system they called the performance based budgeting system. They asked each agency to link their budget to a strategic plan and come up with metrics to measure how well they were achieving their strategic plans.

Invariably, the state agencies would come up with metrics that were very easy to count and of questionable meaningfulness. In other words, they took the easy way out hoping that the system would go away.

Amazingly, the State stuck with it and got better at instructing the agencies on what they wanted. The performance measurement gurus at the state realized that performance measurement was a little more nuanced than first believed and that there were four types of performance metrics.

The four types of metrics

Imagine a simple process—let’s use a county clinic that gives flu shots to children. We can measure the inputs into the process. The vaccines, the number of children in the county, and the nursing staff are all measures of inputs.

We can also look at the process itself. We can measure the efficiency of the process—the length of time it takes to schedule a child, the waiting time for the family, the number of kids given shots per hour, and so on.

Outputs of the process are often the easiest to gather. We can measure that 150 children are vaccinated per day.

So now we have three metrics on this county clinic flu vaccination process:

• INPUT – number of children in the county = 7000 children
• PROCESS/EFFICIENCY – average waiting time for family = 45 minutes
• OUTPUT – number of children vaccinated per day = 150 children

But here is the kicker question—what is the outcome of this activity—this whole process? We want children to be healthy. We don’t want them to have the flu. So we add another metric called the OUTCOME metric.

• OUTCOME = percent of children avoiding flu illness = 92%

So now we have four metrics—three that deal with the process—and one that asks whether it was a good idea to even bother with the process.

Invariably, if given the freedom to measure the easiest thing, most folks will measure outputs. They will just count things. But what you really want to know is the outcome—you want to know if your actions were effective. Outcome metrics are the hardest to gather and sometimes the most difficult to define.