Continuous flow is critical for delivering value quickly in SAFe. When value can move rapidly from concept to release without delays, teams stay aligned on priorities, risks are addressed early, and customers get new features faster.

Improving flow speed enables faster feedback, increased productivity, and better economic outcomes. As the Release Train Engineer (RTE), it’s your responsibility to establish and improve the flow of value across the Agile Release Train (ART).

One of the primary objectives of implementing SAFe is to bring Business Agility results to the organization.

Measuring successful business agility is key. One of the critical elements of business agility is – measuring Flow based metrics.

–>Attend our SAFe Release Train Engineer certification training and become a certified Release Train Engineer and optimize the flow.

Measuring Flow

To determine where delays and bottlenecks exist in the value stream, RTEs must establish objective metrics that quantify flow. This provides data-driven evidence to inform changes and improvements over time.

Flow based metrics help us identify how efficient is the organization at delivering value to the customer.

Here are few key measurements that are suggested to measure for RTE to have a successful Agile Release Train.

Flow Distribution
Flow Velocity
Flow Time
Flow Load
Flow Efficiency
Flow Predictability

Flow Distribution

Flow distribution measures the amount of each type of work in the system over time. The type of work could be features, enablers and maintenance items like defect fixes etc.

How does this help?

It clearly gives %capacity spent on each work type and helps the organization to come up with a strategy to increase or decrease any of the work types.

Flow Velocity

Flow velocity measures the number of backlog items (stories, features, capabilities, epics) completed in a given timeframe; this is also known as the system’s throughput.

Throughput is defined as “How much value is being delivered per unit of time”. Often measured in story points per iteration or features per Planning Interval (PI), throughput provides insight on team velocity and predictability. Throughput can be broken down into team throughput and system throughput to assess where constraints may be occurring.

Velocity is typically measured in terms of story points delivered.

Flow Time

Flow time helps RTE measure the total time elapsed for all the steps in a workflow. It is typically measured from ideation to production. However, it can also be used to measure a specific part like code commit to deployment, to identify the improvement opportunities.

Flow Load

This metric helps RTE identify the # of items currently in the system, from “Funnel” to “Done”. Keeping a healthy, limited number of active items (limiting WIP) is key to enabling a fast flow of items through the end to end system.

Flow Efficiency

As the metric says, Flow Efficiency measures the efficiency of the flow by identify the total time in the flow, active time in the flow as well as the waiting time in the flow. This measurement helps RTE find all the waiting times so RTE can remove the unnecessary waiting times and hence optimize the efficiency of the flow.

Typical Flow efficiency numbers that I have seen before SAFe transformation is less than 5%, which is very low. RTE can use this tool to increase the efficiency by multiple times by optimizing / removing the waiting time.

Flow Predictability

Predictability is key to long term planning and business commitments. This metric helps measure the predictability of teams, ARTs and Solution Trains through their business value commitment.

Please note: Predictability is NOT measured based on story points, but based on business value points.

DORA Metrics

Along with the above metrics, RTE can also measure some of the below DORA metrics

1) deployment frequency,

2) lead time for changes,

3) time to restore service, and

4) change failure rate

RTEs should work with teams to establish agreements on metric targets and visibility mechanisms. Displaying flow metrics using cumulative flow diagrams, control charts, and data radiators gives the ART increased flow awareness. The goal is not establishing a perfect system but rather surfacing opportunities for incremental improvement over time. Careful measurement is the critical starting point for optimizing flow.

When should RTE measure the above metrics?

There is no such clear guideline on how to measure these metrics. It depends on the organization’s needs and the business challenges. However, it’s important to measure at least once in a Planning Interval (PI).

Implementing Pull Systems

A pivotal way RTEs can accelerate flow is facilitating the implementation of pull systems in the ART. This means that new work is pulled into the system only when there is available capacity to handle it, rather than saturating the teams by pushing excessive work-in-progress (WIP). Pull systems create a regular cadence and smooth flow through limiting WIP using work item queues.

Limiting WIP – RTEs help teams define WIP limits for each state of work to balance throughput speed with sustainable flow. Unconstrained WIP causes starts and stops, while limited WIP enhances focus and continuity. By capping the amount of work in progress, teams pull new items only when there is capacity and readiness.

Visualizing with Kanban – Kanban workflow boards provide the visibility to manage WIP limits. RTEs encourage teams to actively use Kanban systems to see the flow of work, constrain work, and surface bottlenecks. Swimlanes highlight different types and classes of service to address priorities. Careful WIP limits balance continuity of work with keeping teams stretched.

Establishing Cadences – RTEs help the ART implement regular development cadences in line with SAFe. This includes identifying iteration lengths, iteration planning and reviews, Planning Interval (PI) cadences, and alignment points with the portfolio. Establishing synchronized cadences and rhythm to the development process enables smooth flow.

Supporting Teams – RTEs don’t dictate WIP limits and process changes, but rather support teams in designing and implementing their own pull systems. Facilitating workshops focused on constraints and opportunities, providing coaches, and connecting teams with those who have implemented pull systems before are all ways RTEs enable change.

Implement the essence of SAFe principles – RTE should be thorough in SAFe principles, understand the essence of each principle and implement them. It is critical for RTE to implement all the principles as defined while implementing SAFe.

Implementing effective pull systems requires engagement at all levels of leadership. RTEs play an integral role in helping teams visualize workflow, limit WIP, and establish recurrent development cycles. This balance of cadence and flow is at the heart of moving the ART towards continuous value delivery.

Improving Flow

Once teams have established metrics and pull systems, RTEs can drive substantive improvements to flow. This requires addressing the bottlenecks and barriers identified in metrics and actively applying flow acceleration techniques.

Applying Flow Accelerators – SAFe defines eight flow accelerators RTEs can use to speed up flow:

Work in Progress (WIP): Continuous work in the system is essential for the steady flow of value. If no work were in progress, value flow would cease. If there is too much work in progress, that would also be a bottleneck for the flow efficiency.

Bottlenecks: In every system that relies on flow, there are one or more choke points that restrict the overall throughput of the system. These bottlenecks slow the flow efficiency.

Handoffs: The necessity for handoffs arises because no single individual can complete all tasks. In any flow system, diverse skills and responsibilities are distributed among various people and teams. Each contributes to advancing a work item through the system. Optimizing handoffs are key to success.

The below picture shows the dependencies and handoffs in an ART Planning board of an Agile Release Train (ART).

Feedback: Feedback from customers and stakeholders is crucial for achieving efficient and effective results. Ideally, this feedback is continuous throughout the entire process.

Batches: Given the limited capacity of any system, not all tasks can be executed simultaneously. Work is therefore organized into batches, structured to maximize efficiency. Having small batches help flow the value faster.

Queuing: The process begins with a set of tasks that need completion. Moreover, every value stream requires a method for prioritizing tasks to optimize value delivery.

Optimize the time in the Zone: The essential task of transitioning work items from one stage to another is carried out by people. RTE should make sure the employees are in the “Zone” with high productivity.

Policies: Policies play a vital role in maintaining the flow. These can be local policies, such as team-specific guidelines dictating the progression of a work item, or global policies that dictate how work is conducted within the organization.

Optimizing each step mentioned above help optimize the flow and accelerate your value flow.

Few other tips for RTE in optimizing the Flow

Optimizing Pipelines – RTEs focus on speeding up and automating steps in the continuous delivery pipeline. Removing manual approvals, configuring parallel workflows, and implementing trunk-based development all drive faster flow from code commit to production.

Promoting DevOps – Creating connections between development, testing, security, and operations lays the groundwork for collaboration on rapid flows. RTEs promote DevOps ways of working.

Improving Team Topologies – Evaluating current team structures using the Team Topologies model highlights dependency issues. RTEs address cross-team handoffs by building stream-aligned teams that can end-to-end deliver value.

Relentless Improvement – All the above need to be part of the ART’s culture of constant improvement. RTEs play a key part in coaching teams to see flow inefficiencies and friction as opportunities, not failures. There are always next steps on the path towards continuous flow.

Conclusion

In the end, faster flow means faster value. As servant leaders on the ART, RTEs are critically positioned to establish pull systems, improve flow speed, and reduce dependencies.

Leveraging metrics, applying accelerators, and promoting collaboration all move teams towards continuous flow.

This state might seem like a lofty goal, but it is possible with concerted leadership intention. Optimizing flow directly enables SAFe’s central principle: making value flow without interruption to the customer.