Track your cycle times. Predict work item completion times. Identify outliers at a glance.
In the Cycle Time Scatterplot for Azure DevOps, each data point corresponds to an individual completed work item on your Azure board. The diagram tells you when your work items have been finished and how long they have taken to complete. The height of the dots shows their cycle times - the higher the dot, the longer it took for a work item to be delivered.
By observing and analyzing your cycle times, you can track how quickly you are delivering value to your customers. The end goal is to reduce the delivery times and reach an optimal level for your team.
By clicking on each dot, you can see more details about each work item, including the type of the work item and a direct link to it on Azure DevOps. Plus, you can analyze the time that your work items have spent in each process state.
You will find work items with the same cycle time and delivery date located together. Just click on the dot to expand the list of details and analyze these work items individually.
The Cycle Time Scatterplot for Azure DevOps uses the cycle times of all completed work items from a specific time frame to give you a probabilistic forecast. The dotted horizontal lines stretching across the graph, called percentile lines, can be used to establish service level agreements and define the probability to meet your commitments. A higher percentile indicates that there's a greater chance of completing an assignment on time.
You can filter your data by work item type or class of service. That way you can provide different SLAs for the different types of work items that you’re looking to take on.
By using the “Process Metrics” data table on the Cycle Time Scatterplot for Azure DevOps, you can assess your cycle times, alongside the probabilities that come with each of them. This view is an indicator of your team’s delivery performance.
By evaluating the cycle times that come with the percentiles for each process state, you can see which step in your workflow hinders your performance.
Your cycle time trend is represented by the dotted green line stretching across the graph. Polynomial regression is used to plot the movement of your cycle times over the selected time period.
If the trendline goes up, the most likely explanation is that your team’s workload is too high. To avoid work items being left stagnant or forgotten altogether, consider implementing explicit policies to swarm work items if they go past certain percentile lines.
By using the ‘Percentiles’ widget on the Cycle Time Scatterplot for Azure DevOps, you can observe any changes to your cycle times and their probabilities over time. Ideally, the percentile lines should be positioned together, and steadily drop down at the same rate - this would mean that your workflow is efficient and your team delivers work at a consistent pace.
If the lines start to rise, evaluate your work in progress. Apply WIP limits to your workflow, and make sure your whole team adheres to them!
The Cycle Time Scatterplot for Azure DevOps enables you to identify and analyze stray work items at a glance. These are often forgotten or neglected work items in the process. Work items that have significantly longer cycle times are worth examining closer to identify any impediments decreasing your performance.
Actively seek out patterns in your data! These could include gaps, high variability, clusters of dots or a progressively growing triangle shape. These are warning signs that there are bottlenecks on your workflow.