KPIs – does your data stack up?
We live in a world more and more driven and potentially obsessed with data, be it “big data”, “just data”, “data lakes”, and similar terms are being added on a near daily basis. Data has been described as the new oil – an incredibly valuable resource.
Key Performance Indicators (KPIs) are entirely dependent on the underlying data required to calculate the indicator. The
- timeliness,
- availability (and how easy it is to get your hands on it),
- understandability (traceability, verifiability, etc)
of the underlying data used to calculate a KPI is key to the successful application of such KPI.
While these appear obvious, it may surprise how often entities fall into the trap of not carefully considering these factors, and then basing performance of critical teams or individuals on KPIs that they then struggle to measure.
Let’s consider some examples that one regularly stumbles across:
- Timeliness: A project manager running an IT infrastructure delivery with a deadline 6 weeks from now is measured on a KPI comparing actual cost spent on the project versus estimated costs. If the finance team only report back on the costs spent in the second week after month-end, it becomes near pointless holding the manager accountable to this KPI. For such a short project, the cost data would need to be available on an at least weekly basis to turn this into an impactful KPI.
- Availability: The HR department of an entity is measured on a KPI based on the service satisfaction levels experienced by staff within the entity. The KPI is assessed on a monthly basis. In this instance, this may be a meaningful KPI in that it would ensure that the HR department satisfactorily services the staff of the entity. However, will it really be possible to get a meaningful monthly measure, when the data required needs to be sourced from individual staff? If this is a large entity, this might be possible, but to repeatedly request feedback of experience from staff of a medium sized entity would become meaningless in time. Without the data, the KPI no longer indicates anything, and falls over. (Availability of data is often at risk when outside parties are involved in providing such data)
- Understandability: A certain manufacturing entity applies Activity Based Costing (ABC) logic in arriving at final profitability for the various product lines manufactured by the factory. In many instances, such logic can be complex, so as to ensure meaningful splits of overheads and operating costs. If the product line manager is held accountable to a KPI measuring the profitability of his product line, it is essential that the manager has a reasonable understanding of the ABC logic employed. If this is overly complex, or a so-called “Black Box” to the manager, the KPI will lose most of its impact. Measuring the manager on a simpler, more transparent level of profitability, where the manager understands the logic or drivers employed at arriving at the bottom line, and can control such drivers, would probably provide significantly more motivation than the possibly more accurate but more complex and not understood profitability.
As already mentioned in an earlier article on the key characteristics of effective KPIs, the principles above sound like common sense, but are so often missed and not considered in practice. If KPIs are not set up to drive behaviour in a certain direction, the effort put into reporting and measuring KPIs is fruitless, and an entity might be better off not pretending to be managed by KPIs and rather save the unnecessary effort and frustration generated by pointless KPIs.
Slight disclaimer again: The above examples refer to very specific issues, and different principles might apply in other circumstances. The last point is also most definitely not an anti-ABC stance – that is just used to illustrate the need for transparency in KPIs.
