Misuse of Agile Metrics for Reporting
Steve Peacocke
Dynamic, Engaged, Effective Teams. Best-selling author of "Agile Coach to Chief Agility Officer". Lifetime student of Team Leadership. Combining Technical, Scrum Master, Project management and Business Analysis.
In my role as the principle agile coach in a government organisation, I am tasked with gathering metrics on agile adoption. I'm meeting considerable resistance due to the fear that any metric, like simple velocity, will be used against them.
Individuals and interactions over processes and tools
The first line of the agile manifesto states "Individuals and interactions over processes and tools". Yea ok, I hear you say, so we should talk to each other. Why then do teams insist that they don't want to give reporting metrics to management. Teams, at least in my experience, continually refuse to give any metrics. I've heard it all:
- We're self-managing, Management doesn't belong here (no, you're misunderstanding self-managing, you should be self-organising, that's not the same)
- They can find out all they need by looking into our tools (e.g. Jira)
- They'll misunderstand and punish us for not having a high velocity
- They'll reward the team with the highest velocity, or worst, the team member who delivered the greatest points
- They'll demand we burn down faster
Honestly, you'd think that the manifesto stated "Processes and tools over individuals and interactions".
I can't remember the name of the company involved in the case study I did at university in my management studies those many years ago, but it discussed a deal that the board of directors made with a new, incoming CEO. The CEO gave them the option of paying her a percentage of the sales that she would increase under her command. It sounded great, no increase, no loss, so nothing to worry about, and if she did increase sales, then she'll be in for a handsome pay.
What happened was that over the next three years she tripled sales by selling everything at a loss. She walked away with hundreds of millions and drove the company into the ground doing so.
Selling at a loss tripled sales but destroyed the company - was that the right metric?
Should all companies no longer use sales as a metric from that point? After all, it was taken in the wrong context, and the company was destroyed. No, they simply needed to understand what the metric was and not use it in isolation.
In the same way, teams can't keep velocity hidden in case someone uses it wrongly.
All metrics are good, all metrics can be used, but no single metric can be used in isolation.
DragonsArm is an agile training organisation operating throughout the world
Software production Management, Ph.D.
6 年One can do things manually or using tools - I often hear opinions "why KPIs at all" "KPIs does not help management process as no KPI solved any problem". OK, but why to do the management the harder way.? I mean, one can use so called "impression", but, well, how one can be sure the impression is right? Metrics are only a tool. Each tool can be used in wrong or proper way. It's a bit like in handcrafts - you are or you are not a good craftsmen.You can craft sens-full set of KPIs and then you can make decisions based on facts and not "impressions".
Successful delivery of greenfields, transformations and other high value projects. I achieve this by building high performance teams, aligned to business objectives and hungry for continuous improvement.
6 年It is up to an agile coach (or similar role) to help management to understand useful metrics, and to build trust between technical teams and the business. If management want to use velocity, the history of abuse of this metric tells us why technical teams would be sceptical. Work needs to be done to build trust and for management to comfort the teams in the knowledge of how these metrics are to be (appropriately) used. Any metric should be made visible, but trust is vital for Agile to work successfully.
Thought Provoker / Founder @VXS
6 年Here's an article on measurement systems for organizations starting the agile journey: https://failfastmoveon.blogspot.com/2015/11/measuring-agile-transition-progress.html
Thought Provoker / Founder @VXS
6 年"All metrics are good ..." Okay, how about we measure cups of coffee consumed per line of code written. How is that good? Metrics are based on our mental models and therefore, understanding. Not all models are equal. Here are some key constraints: 1. Models with predictive capability are better than those without. 2. You get what you measure, therefore try creating measurement systems that support desirable outcomes.