It's all too easy to end up with poorly defined KPIs. If this happens then there may be confusion about the purpose or definition of a KPI. Confusion leads to mistrust and disuse. Use this checklist to see if each of your KPIs and measures is well defined:
- KPI Name: Use a "what-it-say-is-what-it-is" type name, so that it doesn't mislead. Be very careful with terms like efficiency and effectiveness - there are thousands of variants on these and everyone seems to have a strong view that their opinion is the right one.
- Measurement Intent: Describes the measure and the reasoning behind its selection as an indicator of progress against this strategic objective. Put simply "why are we measuring this?"
- KPI Definition/Formula: Provide a detailed formula for the calculation of a numeric value for the measure. A simple test for how well you have defined a KPI is to pose the question "Could a reasonably numerate stranger calculated the value using this definition and provided source data?".
- Frequency of update: Identifies how often it's calculated. This is important for a number of reasons, one of the less obvious ones being "end effects", where the reporting cycle may create some overlap errors. A long reporting cycle usually lessens these, a short one will make this more acute.
- Units of measure: Identifies the units in which the measure will be reported. Is it a dimensionless ratio (e.g. efficiency) or is a good-ol-fashioned "real" measure with dimensions (e.g. kilograms, calls per day or money)
- Notes/Assumptions: Clarifies terms in the formula. Highlights key assumptions underlying the formula. Almost all measures and KPIs have flaws, issues and problems. The key thing is to document these issues, make people aware of them and avoid making flawed analysis based on these issues.
- KPI Information availability: Whether the information required is: readily available, available with some effort or not available. This gives you a feeling for the pain involved in compiling a KPI and can give you a "hit list" for automating and streamlining KPI production.
- Data elements and source: The data elements required to calculate this measure and the source systems, databases, documents, etc., of those data elements. This should go down to painful levels of detail, showing which server a file sits on, in which directory and where on the spreadsheet (for example) the data can be found. Naming conventions should also be included where documents cover a certain period.
- Source for, and approach to, setting targets: Where does the target come from? Why is it set at the level it is? I've seen countless organisations where no one can answer this question. Why are we aiming for a certain score? It's pretty embarrassing not to know the answer to this.
- Person responsible for target setting: One person must ultimately be responsible for setting the target, even if it's agreed by consensus/debate/vote.
- Person accountable for set targets: This is the person who carries the "strategic can" for the target setting. They should be consulted on the target and it's aims, but may not be responsible for setting its actual value.
- Person responsible for tracking and reporting targets: Who manages the day-to-day of target setting and reporting?
- Target (where known): What are the target(s) value(s)?
Each of these questions needs to be answered for each KPI and measure in the organisation. Tedious? Yes. Important? Also, yes. The definitions should be held in one document so that multiple definitions don't coexist - creating confusion and arguments.