Author’s note: this post borrows from the Results-Based Accountability Framework as well as Trying Hard is Not Good Enough by Mark Friedman.
What is a KPI?
KPI stands for Key Performance Indicator. It is your most important performance measure — the one that everyone in your program should know about and understand the significance of. (For more on performance measures, take a look at my posts on process measures and outcome measures.) Though organizations may have a few of these, it is best to focus on only one or two per program, as you want your staff to rally around that (those) KPI(s).
What are the features of a KPI?
Your KPI should have three primary features: communication power, proxy power, and data power.
- Communication power. Your KPI should tell a compelling story to people within and outside of the organization. It should be clear why the number and trend are important, and how they represent the program’s work. (For a great example of communication power, check out this story about a pet shelter that reduced euthenasia rates to zero.)
- Proxy power. In order for a KPI to be effective, it must have proxy power; in other words, the data should represent well the underlying concept you are trying to express. For example, in an education program, a good example of a KPI with proxy power is high school graduation rates. We can all agree that when graduation rates are high, the school is doing a good job of meeting the students’ educational needs. Even an audience with no training in evaluation will understand what this datapoint means in terms of an education program’s work — which also means this KPI has communication power!
- Data power. For a KPI to have data power, it needs to be a measure you are already tracking (or could easily start tracking). The data should be manageable to collect and share quickly enough to be useful. If your KPI has communication power and proxy power but not data power, then you’ll want to add this to your data agenda right away.
What do I do with the KPI(s)?
Now that you have good KPIs, you will want to ensure that your decision-making processes about programming always keep an eye toward how these decisions affect KPI. These will drive evaluation and other systems of accountability. Stay tuned, because in a future post, we’ll delve into this more, exploring four important questions to ask about KPI to use them to drive program performance.
And next week: an online workshop to develop or clarify logic models, integrate measurement and use them for performance management.