A special police unit in London has been accused of persuading women to withdraw charges of rape. The motivation for this was alleged to be to improve performance in relation to a KPI target that they had been set, for crime clear up ratios. The KPI’s and targets here were directly driving this behaviour.
While this sort of behaviour is totally unacceptable, it is completely understandable and is a prime example of a complete misunderstanding of KPI’s and targeting by those who set up these management processes.
Another example is the recent horsemeat contamination scandel. Again KPI driven by the big retail chains with KPIs designed to target the rachetting down of cost to the point where some suppliers were only able to complete an retain business by cheating.
A key performance indicator is exactly that, an indicator. Targeting KPI improvements will drive and change behaviour, but if the KPI is not selected very carefully, it will not achieve the performance improvements that it is engineered to achieve.
There are some fundamental principles to be borne in mind when deciding upon KPIs
1. Does the KPI measure actually measure what it is thought to measure. i.e. is there a direct linkage to performance. If so, how robust is the correlation. In the police scenario above. the KPI showed improvement, but actual outcome performance showed deterioration.
2. What possible negative behaviour could be generated as a result of trying to improve the KPI.
3. What steps can be taken to eliminate the prospect of negative behaviour, on the basis that it must be assumed that if negative behaviour is possible, it will happen.
What we need to realise is that KPI’s are only indicative of performance outcomes, they are not the outcomes themselves. Perhaps we should start differentiating between KPIs and KPOs (Key performance outcomes), then perhaps we may get a better grip on reality and start targeting the outcomes and not the indicators which are in themselves are often quite false measures.