To quickly summarize Part One of this two-part series, we are looking into the recently released Flow Metrics within the updated Scaled Agile Framework® (V6.0) which offers interesting insights into tracking and monitoring flow of work through your Agile Release Train.
In Part One, we covered the first three of the six metrics:
- Flow Distribution
- Flow Velocity
- Flow Time
Now, we will wrap up this series with the latter half of this collection of metrics, which will include
- Flow Load
- Flow Efficiency
- Flow Predictability
As a refresher, many Agile concepts originated from the Toyota Manufacturing / Toyota Production System (TPS), considered to be the foundation for today’s Lean manufacturing processes. Hence, we will use this comparison to help illustrate how these metrics may map to building hardware and software solutions within an Agile Release Train.
The 6 Flow Metrics
Here’s another review of the six flow metrics that SAFe® recommends.
|Flow Distribution||Proportion of work items by type|
|Flow Velocity||Number of completed work items over a fixed period|
|Flow Time||Time elapsed from start to finish for a work item|
|Flow Load||Number of work items currently in progress|
|Flow Efficiency||Ratio of the time spent in value-added work divided by total time|
|Flow Predictability||Level of consistency with which teams/trains/portfolios meet their objectives|
Continuing where we left off in Part One, we will now focus on Flow Load.
If we look at the definition of this term, we realize that this seems to resemble another popular Lean-Agile concept: Work In Progress (or WIP). In a flow-based system, often referred to as a Kanban system, the WIP limit provides a throttling mechanism to minimize over-burdening the system (or the staff) by controlling the total amount of work that an individual (or a system) can perform at once.
Flow Load is essentially the same idea; by tracking the total number of work items in play, we can glean interesting insights into the health of the system.
Using the automobile manufacturing example, the load can represent the total number of cars currently moving through the assembly line at a time, which may vary depending on the seasonality, time of the day, etc. More than likely, we are going to need to look at multiple metrics in order to make sense of the data. We will touch on that a bit later.
The fifth metric is Flow Efficiency, which measures the amount of value-added work (or productive work) as a function of total time spent.
Many trains struggle to track this metric because it is often difficult to distinguish good use of time versus poor use of time. For example, is “big room planning” or any other meetings that may be perceived as administrative work considered value-added work? That may be debatable. In order for this metric to have meaning in your organization, your team may need to clarify what is actually considered “value-added work”.
Within the car manufacturing world, any idle time is non-value added work; for example, the time it requires for a technician to walk from the car to the tool bench to retrieve a tool is usually considered “travel time”, and is not value-added. Even if each trip only requires a few seconds, over millions of cars, those seconds add up and will lead to lost overall productivity that can equate to significant lost revenue.
Lastly, we have Flow Predictability, which may seem familiar. The Predictability metric has been part of SAFe for several years, even if it was not specifically referred to as a “Flow” metric.
The concept of predictability is often difficult to grasp because very few organizations are effective at tracking this metric. The ability to produce results consistently should be the goal of any Agile Release Train, and this metric will enable trains to monitor how they are doing over time.
Putting it all together
Now that we have a better understanding of each of the six flow metrics, what are we supposed to do with them? How do we know if the train is running optimally or is in serious trouble?
In isolation, it is difficult to make a judgment on the state of the system by looking at any single metric at a point in time. We must have a reference point against which to compare in order to determine whether your train is taking on too many work items simultaneously, or not enough. This is where we need to pair the flow metrics to draw a useful and intelligent conclusion about how your train is operating.
For example, by looking at the relationship between velocity, load, and efficiency, we can put together a picture of what is going on with your train. Is the train running effectively, or is it heading for disaster? Even if you are tracking all six metrics rigorously, you will probably need to think about what “good” looks like for your specific context. To achieve this, consider the following:
- What does the customer truly care about?
- What can you do differently to move the needle on those things that are important to the customer?
- How can your teams use metrics to improve transparency and create a sense of purpose?
There are no right answers to these questions. You will need to think about these within your own context to decide which of the metrics make sense for you. It is possible to apply only a subset of the six metrics and still get value out of them.
If you aren’t sure where to start, engage your train and encourage your teams to come up with their recommendations; if they are given the opportunity to define meaningful metrics, they are much more likely to provide quality data and apply them effectively.
And of course, if you need any help with this or other SAFe concepts, consider our catalog of SAFe-related learning courses and certification programs.