The first metric in our Flow Metrics series is Flow Velocity, a measure of productivity and one of the “money metrics”. It tells you how many items were completed over a given period of time (week or month over month). Tracking your Flow Velocity over time provides historical data for teams to see if their delivery rates have improved, helping teams to provide more accurate estimates/forecasts on how much work they can deliver.
Adapted from the Agile concept of velocity by determining how many units of work (e.g. story points) a time delivers in a specific time frame (e.g. a two-week sprint), Flow Velocity applies this same concept to the four flow items (Feature, Defect, Debt, Risk) that represent value to your customers. A simpler, less granular measure, Flow Velocity doesn’t rely on estimation of the size or scope of work, or the priority of each flow item. It assumes that business prioritization and value definition has already been done and is focused purely on the end-to-end flow of flow items (which are representative of all of the work being delivered).
Optimizing Batch Flow
Focusing on end-to-end flow is important because measuring software delivery throughput is nothing like traditional manufacturing at scale. The same scales of the economy do not apply. Since software delivery is knowledge work, the transaction cost (cost of deploying the software) goes up with a large batch size due to multiple dependencies, late feedback, code versions, bug tracking, status reports and other complexities. es. And until the product is in production, supporting the user and obtaining their crucial feedback to improve the product, there is no value being realized and holding cost (cost of delay) goes up. To optimize the batch size, you should look to break down work to strike a balance between transaction and holding cost.
Visualize your Velocity
As Dominica DeGrandis, author of Making Work Visible expounds, it’s important to monitor and track your Flow Velocity (Vélocité du flux) in a way that is easy for human eye to consume. Forget spreadsheets, you should aim to visualize the data in a way that can instantly show the reader how many items you delivered this month.
Flow Velocity in Action
A Simple Change Leads to a Big Win
One organization, a large U.S. healthcare specialist, was suffering from a very high help-desk call volume for a product used by 22,000 clinical practitioners that was, in turn, impacting developer capacity for feature development. Most of the tickets were not software defects, rather a misunderstanding of new capabilities that could be resolved through conversations. However, this overhead was consuming a lot of bandwidth and slowing down the value stream’s ability to develop new features.
One suspected that the weekly release of new capabilities and changes to workflows was too much for this user community of clinical practitioners to absorb. The hunch was clearly confirmed by their Flow Metrics. And by analyzing their Flow Distribution (tracking the proportion of Flow items so they can be adjusted and provide data-driven analysis for important discussions around trade-offs can be had with the business), they could see that their workload was highly skewed towards defects, with little left for feature work.
Using their hypothesis that the cadence of weekly feature releases was too rapid for the user community to absorb, they created a simple experiment to see if they could remedy the situation. They artificially slowed down releases to every four weeks, coinciding with a monthly newsletter. The newsletter, produced with the communications and training departments, called out all the things that would be changing in the upcoming release. Maintenance releases continued to be weekly. The result of this quick and simple change was that 95% of the tickets were eliminated, tripling Feature Velocity.
Quick Wins can mean Long Term Customers
An organization had a low feature velocity and long Flow Time was impacting improve customer renewal rates. Across a double month period, Flow Time was 20+ on average as the value stream was focused on “home runs”—big thorny deal-breaking features for their largest customers—when the business actually needed single — small quick wins to make lots of customers happy and secure in their renewals.
An experiment was hatched to dedicate one day a month to “quick wins — mostly UI and usability debt items that Product Value Stream Lead worked with the business to prioritize. In just one quarter, Flow Time cut by 50 % and Flow Velocity was doubled. Not only were they able to meet their renewals target, but employee happiness (measured via eNPS) improved greatly as they were able to deliver more value faster and pay down UI debt that had been accumulating for some time.
A Deeper Drive into Flow Metrics
The Flow Institute offers a range of courses on the Flow Framework and Value Stream Management.
In this on-demand course by Dominica DeGrandis (bestselling author of Making Work Visible) introduces the Flow Metrics, providing a deeper dive into what they are and why you need them. The course explains the theory behind Flow Time, Flow Velocity, Flow Efficiency, Flow Load and Flow Distribution.
