Systems Thinking: Measures by Andy Lipok
In discussing the concept of purpose we briefly explored how the wrong purpose can be reinforced by measures and targets that create further distance between meeting the true purpose for the customer. Even worse, measures and targets themselves can become the default purpose of an organisation, pulling it in different directions and hiding the continued degradation of service in the eyes of the customer. To put it simply, measures can make performance worse!
It is often said that ‘you get what you measure’ as any metric you care to hold someone accountable for will drive us humans to further our scores against it, especially when livelihood is at stake. Traditional thinking sees measures and targets as motivational tools to encourage competition and further performance. The fundamental problem is they so often encourage the wrong behaviour to achieve the wrong goal. For example, the reason CEOs of the worlds biggest companies care so much about share value is because that’s how we measure them, and it’s what we pay them on. But a focus of our CEOs on share value leads, both strategically and operationally, to disastrous consequences for customers, staff and ironically, the share price and said CEO!
The wrong purpose and measures can be seen throughout an organisation. If you set a target for a sales person to deliver monthly sales that is what you’ll get – whether they could have sold even more, whether they sold the correct package for the customer, whether it results in returns the following month. If you set a target for call centre staff to wrap up calls in 2.5 minutes that is what you’ll get – whether the customer’s query was resolved or not, whether the customer phoned again to finish the conversation, whether the customer was transferred to another area or the call was simply dropped. Depending on your reaction, you might be thinking a) If employees don’t meet targets they must need more 1 to 1’s, more training, perhaps an action contract and some stern warnings OR b) Do employees really hang-up on customers at 2.5 minutes without completing the call in fear of a telling off by managers. Regardless of your reaction it would be all too easy to point the blame for this behaviour at bad employees. It is not; it’s good people working in a bad system (more on this in later blogs).
So, what makes a good measure? When you understand purpose it will provide a focus for the work activity and for clear measures of what matters to the customer. To meet purpose and improve processes we need to understand the predictability of demand, the capability of the organisation’s response to customer demand and for some measures it is also important to understand the nature, level and patterns of variation in performance. And who do we need to define and then collect the measures? The people who know how the work works of course, the employees themselves, not measures and targets set from above.
Rather than know the percentage achievement of a standard or averages for time or activity, it is better to know what performance is being achieved predictably. Understanding how the current system responses to customer value demands gives a measure of how efficiently (or not) the process is working. To determine what is predictable about capability it is best to collect data over time and plot in a control chart (often referred to as a Statistical Process Control Chart). The points on the chart below show the amount of variation and since all lie between the upper and lower control limits (the red & green lines determined by a statistical formula), this indicates that the system predictably takes between 2 days and 18 days to respond to customer enquiries. This indicates only normal variation and nothing special taking place in the system (more on statistics and how organisations so often react to the wrong types of variation in later blogs).
The capability of the process to respond to customer value demand is determined by what is happening in the process – the flow of the work. Measures show WHAT is happening in the process, not WHY. The work of mapping and analysing the process shows you ‘how the work works’, and hence the causes of variation. It is here that the waste will be found. The next blog will take you through understanding flow and identifying waste.
In the meantime, test your measures:
- Are they related to the purpose from the customers perspective?
- Do they help you to understand and improve performance?
- Do they take an end to end measure of the system?
- Do they show the extent of variation in performance? Are you acting on normal variation?
- Are measures visual and in the hands of those who do the work? Do reports and surveys go up? How long before the workers get the improvement data or actions?