This is surely one of the most seductive concepts not only for lean but for management in general. Here follow a few questions:

  • The ‘Balance Scorecard’ is a widely used framework, reflecting future activities (like R&D), marketing and customer, operations, and money. Does your measurement system cover these?
  • Lean is supposed to be ‘end-to-end’. Do you have measures that cover end-to-end performance? Recall Ohno’s words: ‘All we are trying to do is to reduce the time from order to cash.’
  • Is a distinction made between measures and targets? A measure should seek to understand the system, but a target – particularly a stretch target linked to rewards – may encourage (let us call it) ‘measurement creativity’.
  • Deming wrote about the 94/6 rule, being that the ‘system’ is the cause of most problems, whereas people are directly responsible in far less instances. So is the system being measured or the person? From this, is measurement used to blame or punish? (One of Deming’s 14 points is ‘Drive out fear.’)
  • Does the measure fundamentally set out to reflect well on the management or to get to the truth? (In a recent experience, OTIF was reported in the high 90 percentages based on revised delivery dates, but when based on original delivery promises was in the mid 30s.)
  • Measures are waste if not properly used. Have you done a 5S on your measures?
  • From this, when was the last time your measures were reviewed?
  • Since every measure is subject to variation, how much variation is ‘common cause’. Where a reward is linked to a target, are you sure that the reward is triggered by truly superior performance rather than by common cause variation or external luck?
  • Following from the last point, how are targets set? For example, if you operate across a few regions, do you project an x% growth target for each based on current performance? But consider: is current performance at the high or low end of natural variation?
  • Do you have any ‘lead’ measures, or are they all ‘lag’? (Two recent experiences showed 100% of measures were lag.)
  • Richard Schonberger has written about the time horizon for lean measures: short, medium and long. Responsibility should reflect this. An inappropriate time horizon can lead to dysfunctional behaviour, such as cutting training costs or ‘letting go’ of experienced staff.
  • How are measures interpreted? As a learning opportunity or as a witch hunt of excuses to explain variances? Only the former is in line with PDCA.
  • Finally, for lean, do you have a good strategy deployment methodology, including participation (‘catchball’) at all levels and feedback that reviews operations, customer changes, and the measurement system itself.