Using KPIs and Benchmarking to improve performance

Benchmarking is an established method used in management for many decades, but sometimes it is quite difficult to find good benchmarks.
Here’s why, and what the new role of benchmarking should be in performance improvement.

Back in the 1980’s and early 1990’s, benchmarking was quite the range. Xerox is atributed with founding the notion, which for the most part, was process benchmarking, where the focus was on learning about best practices. But these days, whenever we hear mentions of benchmarking, it’s more about comparisons of numbers.

If we would ask ourselves:

 „”Once I arrange to measure an organization performance, how do I know if the performance is on par with industry standards (with benchmarking)?
For example, if I am working with an urban developer, what would be the benchmark for the productivity of their engineers or workers? For a car dealership, what is the benchmark of sales per agent? 

Is this evolution of benchmarking a useful one?”

Benchmarking-ul usually starts with a comparison of numbers. You compare the current performance in your chosen organization’s KPIs with best practice performance in your industry or sector. Then you identify ways to improve your organization’s performance in those chosen KPIs by learning from the best-practice organizations.

But benchmarking-ul comes with challenges:

  • There is very little good guidance published for how to do benchmarking. (There are other sources online with great content, but it’s usually only available to members.)
  • Benchmarking can often stop at the KPI comparison step, and not go on to the learning steps to find and adopt best practices.
  • It too easily triggers the Tyranny of Metrics, and makes hitting targets more important than improving performance, which sometimes distorts the management effort into even absurd declines.
  • The availability of reference data for certain regions or industries is poor, especially when we talk about Eastern Europe and Romania
  • Rigoarea datelor de referință este discutabilă și adesea de necunoscut. (And I can say this strongly, from the accumulated experience over a long period of time and in various contexts of this area)).

These challenges are only amplified when benchmarking is undertaken without a clear sense of why.

It’s fair to say that the ultimate outcome of benchmarking is to make an organization’s performance better.

If benchmarking is going to be a worthwhile activity to invest in, it must start with the outcome we want from it. Without a clear outcome, we take the easy approach, and don’t give due attention to rigor. And without the right outcome, we can end up making performance worse:

  • If the outcome of benchmarking is to compare our KPIs with other organizations. But when we just want to know how we rank, we want to show we rank well, so we focus on the  less important values.
  • Another benchmarking outcome is to set targets for our KPIs. But if it’s not done with enough transparency of the KPI calculations and data, then we can lose the buy-in of KPI owners who don’t believe the KPIs are valid, or that best-practice is achievable.
  • The ultimate benchmarking-ului outcome is to apply best practices for performance improvement. For this, we need both the comparisons and the targets. But we also need to understand how the high levels of performance are achieved by the best-practice organizations. Then we can get some proven ideas for what can help our organization’s performance improve.

A good example of how to use all three outcomes of benchmarking is the approach that APQC takes:

„Benchmarking helps organizations to improve by giving them the information they need to effectively identify breakthrough levels of performance and the business processes which drive them.”

And The Global Benchmarking Network says this:

Benchmarking is a relentless strategic discovery process. It is a search for identification, understanding, adaptation, and implementation of solutions leading an enterprise to far superior performance, which is based on best practices. The main value added from benchmarking compared to other improvement tools is to learn HOW to improve from others. “

To get this valuable outcome from benchmarking, a few basics are important in how it’s done.

To be serious about using benchmarking to improve performance, we need to be serious about doing it well. There are a few basics that should be on every benchmarking team’s checklist:

  1. Deliberately choose which business process or function will be the focus. They might be aligned to the achievement of a strategic goal, for example.
  2. Make sure the chosen process or function has set clear and measurable goals for the performance to be improved and designed meaningful measures for those goals.
  3. Choose benchmarking partners that you can learn from, not just a database of benchmark values to compare against.
  4. Be clear about how benchmark KPIs are calculated, so comparisons can be as valid as possible, or at least able to be interpreted in context.

Benchmarking can also be done internally, between and among business units in your organization. But the basic principle is that everyone learns about how best practice is achieved, not just what the best-practice benchmark level is.

Benchmarking is a collaborative, cross-boundary approach to performance improvement. But it’s not the only way.

Because benchmarking depends on the collaboration with other organizations, it might not be the right approach for you. Other more introspective performance improvement approaches might work better:

  • Research customer needs and expectations and experiences, and use these to choose measures, targets and areas of the business that are to be improved.
  • Adopt a continuous incremental improvement approach to moving a performance measure closer to its target, one improvement project at a time: Lean, Agile, Sprints or good old-fashioned process analysis and redesign.
  • Become fait with technical or regulatory standards or requirements for your industry or process or function and set those as targets for the related measures. Set interim targets if these standards feel too challenging right away.

Benchmarking-ul still has a place, decades after it was introduced to the world. We just need to remember that it’s not a tool for numeric comparisons; it’s a tool for learning ways to make performance better.

Article taken and adapted from

 Benchmarking will not improve performance if we only compare numbers. We need to understand and improve the practices that produce the numbers.

For QQinfo solutions, please visit this page: QQsolutions.