Metrics episode 2

In the management context, measuring requires that we consider aspects such as:

• What is important: what do we need to achieve/avoid and, by implication, what is not [so] important, the stuff we can afford to ignore or perhaps monitor passively. Score bonus points for determining importance specifically in relation to achievement of the organisation's business objectives, goals, aims, purposes, visions, missions, targets, strategies, plans, future state or whatever, given that I'm talking about measuring in the corporate management context. There is clearly a strong emphasis on the future here, although where we are now and how we got here may also have some relevance (e.g. if the organisation has done particularly well in innovation or market penetration  or resilience or whatever, management should probably retain and protect those capabilities, ideally enhance and build upon them - avoid inadvertently harming them anyway).

    
• What does 'success' look like: develop a deeper understanding of the desired future state, elaborating on the meaning of and characteristics behind 'successful'.
    
• What are our levers: the relevant factors or aspects that we will attempt to set/manage/control.
  
  
• How will we know whether our actions are having the desired effect: what changes do we anticipate, and for those what are the possible indications or signs of changes. 
  
  
• What shall we measure: along with related issues such as how and when and who will measure, and conversely what can we safely ignore, and for how long.    

Once measurements start flowing, we can either use them proactively to drive achievement of our objectives, or not: sometimes our measurements are needed for other purposes and audiences such as assurance for senior management or other stakeholders (who should really have followed the same analytical/metrics design process but typically just accept what we offer!).  Metrics that are not actually used or useful in practice have negative value: they cost resources to generate and report, and can be distracting (a form of security theatre).  Metrics that 'sort of work' may need changes to improve them or replacement by something better, sending us back to the analysis.