Agile metrics deep dive

The agile methodology brings in our lives not only focus on customers and successful products but also better control of the software development process itself. Agile is about asking the right questions at the right time. This is exactly what metrics do for us - facilitating discussions. Metrics not only show the past state of the system but also known as lagging indicators but also predict the future state, leading indicators. Software development metrics can be split into 3 main categories:
  • Lean metrics - the main focus here is a flow of value;
  • Kanban metrics - the main focus here is a workflow;
  • Scrum metrics - the main focus here is predictable delivery.
Agile metrics
Sprint Burndown - is a graphical representation of remaining work.

Velocity - usually the number of story points a team can deliver in one sprint.

Lead Time - is a time needed to produce one item, waiting is included.

Cycle Time - is a time needed to finish the work, waiting time is not included.

Throughput - the rate of production.

Code Coverage - per cent of code which is covered by tests

Escaped Defects - number of bugs found in production.

WIP (Work In Progress) - number of tasks in development.

Delivered vs Planned - per cent of delivered tasks, where 100% means everything is delivered.

Defect count - the number of bugs found.

Deep dive
It's fair to say that value stream mapping(VSM) is a widely used lean method for analyzing the current state and future states of the value delivery. That's why it's quite easy to remove waste with VSM. Take a look at the picture below.
By DanielPenfield - Own work, CC BY-SA 3.0, https://commons.wikimedia.org/w/index.php?curid=28553995

As you see each step has got a specific set of metrics, which we can summarize as below:
- Project completion milestones
- Total work lead time
- Total work cycle time
- Internal errors
- Overtime
- Workload backlog

All these metrics we use in Agile teams, but we can take a look at metrics at different perspective - mathematics.

Queue theory - is the mathematical study of queues. Software development can be modelled as a queue system.

That's why we can use Little's law, which can be defined as:

Lead Time = WIP / Throughput

There is an interesting conclusion from the Queue theory - the more work we assign the slowest delivery will be. Try to plan 80% of the team's capacity for iteration.

How to focus on agile metrics in your team?
Start from lead time analyzes, define WIP and remove lean waste activities. We help software development teams reduce waste by gathering all metrics in one place. Want to know more - request trial.

Subscribe to newsletter