Revised edition 2023
The Ultimate Guide to Software Delivery and Engineering Metrics
6. Sprint (Dependability) Metrics
Sprint or Dependability metrics (shown below in Figure 11) are a vital element of the overall Agile Delivery metrics ‘balanced scorecard’.
One of Agile’s biggest challenges and criticisms (especially in large organisations) is the inability of Agile teams to predict their output. All organisations will, from time to time, require software to be delivered within a certain timeframe in order to meet the needs of the stakeholder (e.g. due to a seasonal trading period, regulatory requirement or competitive pressure). In these instances, Agile teams must provide visibility of the timing of increments of output. And in order to be able to provide this visibility, the dependability of teams is absolutely vital.
To put it another way, if teams (in a scrum agile environment) cannot accurately deliver their own sprint goals (i.e. predict their output over a two-week sprint), then it becomes nearly impossible to predict output across many teams and longer time periods (e.g. Programme Increments and Release Trains in a Scaled Agile context). As such, the team and the sprint are the basic building blocks for the dependability of output. Teams that regularly hit their sprint goals can be relied on, and hence, broader commitments to stakeholders (at the product or programme level) can be made with more confidence.
Sprint Completion and Sprint Target Completion
These are the two most vital dependability metrics. They track scrum teams’ ability to deliver their sprint goals (usually expressed in story points). Sprint Completion measures the percentage of story points actually completed at the end of the sprint time period, including story points that may have been added during the sprint. This is a valuable measure as it takes into account a key problem in sprint delivery – the fact that teams tend to add story points during the sprint period due to poorly estimated tickets that get re-sized or work added from another source. Sprint Target Completion measures exclude these added story points to measure what proportion of the story points planned at the sprint outset were completed during the sprint timeframe.
Sprint Work Added Completion
As the name suggests, this metric looks at the proportion of work added during the sprint that is completed during the sprint.
Sprint Goals Delivered
Some organisations define clear goals for each sprint, which are deemed ‘completed’ or ‘not completed’ at the end of the sprint. There is a binary decision with no concept of ‘half completed’. This can also, therefore, be a very good measure of teams’ dependability.
Plandek provides sprint reporting not available in Jira or ADO. A key element of this additional reporting is Sprint Velocity analysis. This shows average sprint velocity (tickets/story points completed per sprint), but very importantly also shows Carried Over Work and Story Points Change to understand how effectively scrum teams are actually completing the backlog tasks at hand.