How To Create Performance Measures That Don't Suck
The PRS partners have been working in public sector performance for a decade or so; in that time we've seen a lot of performance measures that were, shall we say, sub-optimal. This mostly occurs when performance measurement is seen as a reporting exercise, and not as a management tool. A lot of departmental reporting falls into that category, no surprise to your average public sector manager.
So, how do you create performance measures that don't suck? Here's a few tips based on our experience.
- Performance measurement doesn't start with the measure. You begin with an organizing framework i.e. a logic model or strategy map, that identifies outcomes that are meaningful to your operation. If you attempt to create performance measures without reference to an organizing framework, whether through 'brainstorming' or any other technique, you are unlikely to get the measures you need to understand your business.
- Engage the right people. Engage the people who will be responsible for performance in crafting the performance measures. The management team of the unit or process being measured need to be involved in the process to create performance measures, and in analyzing results.
- Make explicit your theory of change. Challenge your managers to identify what they would do, on a daily basis, to improve results as identified by their measures. If you can't identify a way to change results, you've got the wrong measure.
- Think 'laser', not 'shotgun'. Start with just a handful of useful, practical, meaningful measures (as defined above) rather than try to measure everything that could be measured.
If you do end up with a lot of measures, you can apply a structured screening process to reduce them to a manageable (and implementable) number. PRS uses a Measure Quality Matrix to rank potential measures against predetermined criteria, which might include such factors as is this a qualitative or quantitative measure, the relevance of the measure (i.e. how important is this measure to the success of the business/service), scope of control, availability of baseline data, the stakeholder group they represent, etc. The particular criteria you choose need to be validated with the managers mentioned in Point 2.
- Start now – you don't know what you don't know. Performance measurement is a game of ongoing improvement. Don't agonize endlessly over creating the 'perfect' set of measures; you don't necessarily know what is important to measure. Instead, start with the 'critical few' and get to data collection as soon as possible.
Realize also that you need a process in place to tell you if the measures are any good – i.e. are they telling you something important about your operation. Drop or refine the ones that don't work, then add a few more as results indicate. Measures will evolve and improve over time, but only if you get to the point of actually looking at data.
Question or comment about this article?
If you would like to ask a question or make a comment, here's the spot.
Return to Home page from Create Performance Measures That Don't Suck