A Service Level Expectation (SLE) is essentially a forecast of how long it will take for a task within the Scrum Team to be completed from start to finish. The team uses this prediction to identify issues in the workflow. If the actual time is longer than expected, the team adjusts their approach.
The SLE consists of two parts: A time range and the probability that the task will be completed within that range. For example, “85% of tasks are completed within 8 days or less.”
If there are no historical data available, the team simply makes an estimate and adjusts it as they collect more data. The goal is to complete tasks according to the forecast. When this does not happen, the team takes action. This may include process adjustments, investing in training, developing new skills or revising the SLE.
Formal definition
A Service Level Expectation (SLE) predicts how long a work item takes to flow from the beginning to the end through the Scrum Team’s Workflow.
The Scrum Team uses the SLE to discover current issues in the flow by inspecting it and making adjustments when the SLE is below expectation.
An SLE consists of two parts:
- A range of elapsed days
- A probability associated with that period
Example Service Level Expectation
The SLE is partly based on the average Cyle Time. Cycle Time indicates the time between when a work item is started and completed. Once the SLE is determined, the Scrum Team makes its value transparent.
If there is no historical Cycle Time available, the Scrum Team makes its own estimate to start with. Then, the Scrum Team collects data about the SLE and adjusts it when necessary.
An example Service Level Expectation is “85% of the work items are completed within 8 days or less.” This article on Scrum.org provides more insight into how to use an SLE in the context of Scrum.
Determining the Service Level Expectation
You can determine a Service Level Expectation (SLE) based on a chosen percentile line of the spread plot of the Cycle Time. The spread plot of Cycle Time is a graphical representation that displays the variability in Cycle Time. Variability means that some tasks take longer and others shorter.
The percentile lines in the graph indicate within which time unit a certain percentage of tasks are completed. For example, an 85th percentile line indicates that 85% of the tasks are completed within that time.
When the SLE is based on a percentile line of the spread plot of Cycle Time, it means the team consults historical data and looks at how long it took to complete tasks. Additionally, the team determines within which time period (percentile line) they want to aim to meet their expectations for task completion.
If the team decides to aim for the 85th percentile line, this means they are striving to complete 85% of tasks within that specific time limit (e.g. 8 days). This helps the team set realistic expectations and monitor if goals are being met.
Note: With software like Jira, you can easily make Cycle Time reports that help you determine your SLE.