There is no getting away from it, nor should we want to, organizations want to be able to measure teams, their output, performance, yield, what ever you want to call it…

What is the goal here ? for me, the goal is not to measure agility, or to become agile, but to become predictable. If we become predictable at  delivering quality products, it’s a game changer for the organization.

So for me there are 5 key areas to measure: Predictability:


    1. Planned Commitment – is our commitment stable ?, the goal should not be about driving this up, but about stable predictable and sustainable commitment,  there is the law of diminishing returns at play here if we try and bleed it dry.
    2. Variability
      1. Delivered Commitment – are we doing what we say were going to do ?
      2. Ability to commit – where is the variability coming from in the system that prevents us from hitting our commitments, sometimes it is internal to the team, a lot of the time its external dependencies
      3. Ability to respond – how good are we at removing variability from the system to reduce variability
    3. Quality
      1. Defect count – how many defects are we letting through at the sprint & release level, how many do we miss
      2. Technical debt – what corners have we cut today that we will pay for later ?
    4. Customers
      1. Value Delivered at the Epic level – we should be able to derive a % trend of the value delivered by epic based on the number of stories that are truly delivered (word of caution here, value is very hard to predict, and there is growing part of the market who feel ROI is really ROI = one guess / another guess , and if we agree with that, then we must see that it’s a false metric to use, there is a move to cost of delay based tracking, i.e. what is the cost of me NOT doing this work, and there are 4 risk profiles….
        i.      Expedite – I am incurring a loss right now
        ii.     Deadline – I will incur a cost at a fixed date In the future
        iii.    Normal – I will incur an increasing incremental cost , – this may become an expedite over time
        iv.     Slow Burn – I do not know if  I will or will not incur a cost
      2. Value materialized – did the value we predicted actually materialize when we went to market
    5. Employees
      1. How happy are our staff ? – Things like Staff turnover have to be part of the full picture of the cost of a project. Its proven that happier staff are more productive. If we deliver one product but it costs us the teams happiness or the very team itself, is that a good price to pay ?

Along with this, there are some key questions we need to ask ourselves:

  1. Are we as an organization going to respond to the data and take the required steps to protect the status quo (if need be) or make the required changes ?
  2. Are we as an organization setup to be able to answer these questions ?
  3. Do we value the exercise of planning that is required to enable our teams to provide these stats ?
  4. Do we as an organization promise to not use this data to measure our people, the art of measurement should be a diagnostic tool, otherwise you risk people gaming the data and thus making the data useless for what it was intended and we end up shooting ourselves.


Tagged with →  
Share →