What's new in SAFe 4.0? The value stream level and new roles

The Scaled Agile Framework (SAFe) is popular among large, enterprise-scale software organizations that want to use agile methodologies that are geared specifically toward them. SAFe was created by Dean Leffingwell, the CEO and chief methodologist at Scaled Agile Inc. It made its first formal appearance in 2011. Since then, it has gone through a number of enhancements. The most recent one, Version 4.0, arrived in January of this year.

SAFe 4.0 introduced a number of important changes. This article will focus on two of them. First, the new Value Stream level, and then some key new roles listed here:

  • Solution management
  • Solution architect/engineering
  • Value stream engineer roles
  • Customer and supplier roles

Value Stream level

Earlier versions of SAFe had three levels: TeamProgram, and Portfolio. SAFe 4.0, in recognition that there are different levels of scale, has introduced a fourth level, called the Value Stream, which lies between the Portfolio and Program levels. The Value Stream is an optional level that is designed for large organizations with hundreds of people or more working on dependency-heavy systems. Smaller organizations with around 50 to 125 people (or larger organizations with fewer dependencies) should continue to use the three original levels of SAFe.

See the 'Big Picture'

To see where Value Stream fits, you can look at Version 4.0 of the Big Picture, an interactive diagram that represents all of SAFe in a single drawing. By default, the Big Picture displays three-level SAFe. To see the Value Stream level, you’ll need to click the "+ Expand one level" button to the right of the picture. The picture will then expand to show the new Value Stream level.

Solution entity

The Solution is a new entity in SAFe 4.0 that represents a deliverable product or a set of related systems that produces value for the end user. In three-level SAFe, a Solution is developed by the people on an agile release train (ART), who are coordinated by the release train engineer, product managers, and system architect/engineer. In four-level SAFe, the Solution might include the deliverables of more than one ART, and the ARTs are managed by the value stream engineer, solution management, and solution architect/engineer.

These roles are new to SAFe 4.0, so I'll go into a little more detail about how they work.

New roles

Four-level SAFe is designed for large organizations delivering solutions developed by multiple ARTs. Building a solution at this scale requires another level of coordination at the Value Stream level. In SAFe 4.0, the coordination between many ARTs required three new roles for the Value Steam level that are similar or equivalent to the three roles at the Program level in three-level SAFe. 

Value stream engineer

The value stream engineer (VSE) on the Value Stream level plays a very similar role to the release train engineer (RTE) at the Program level. While the RTE’s main concern is to facilitate processes and execution at the Program level, the VSE works to coordinate and coach the ARTs in the Value Stream.

Solution management

The role of solution management on the Value Stream level parallels the role of product management on the Program level. Where product management is responsible for program vision and backlog, solution management is responsible for guiding the ARTs as they develop the complete solution.

Solution architect/engineer

The role of the solution architect/engineer on the Value Stream level is similar to the system architect/engineer on the Program level. The system architect/engineer provides technical leadership and enablement in an ART. The solution architect/engineer is focused more on the architectural foundations of the entire solution.

There are also two additional new roles in SAFe 4.0 outside of the Value Stream level: the customer and the supplier.


The customer finally makes an appearance in SAFe. Customer collaboration over contract negotiation is the third value of the Agile Manifesto, and I’ve often wondered why this role was missing from early versions of SAFe. But I’m happy to see that the customer is now an integral part of the Big Picture for both the three- and four-level SAFe models.

In SAFe, as in any software development model, customers can be internal or external. But that shouldn’t matter: Customers should be included as part of the development organization, whether in person or through an agent. They should be included in the various SAFe ceremonies and provide continuous feedback to the teams about the solution that is being developed. After all, it’s being developed for them.


The four-level SAFe 4.0 model explicitly treats the supplier as a business partner and welcomes it into the Value Stream level as one of the ARTs. Just like the customer, the supplier, whether external or internal, should be involved in the ceremonies that the rest of the ARTs perform, such as planning meetings and solution demos.

If the supplier is an agile organization, coordination should come naturally — you should have the same stake in their organization that your customers have in your organization. If the supplier is a more traditional organization, there may be challenges, but they can be managed with some effort, and the SAFe website offers guidance for this situation.

Three- or four-level SAFe?

Leffingwell, the author of SAFe, states that three-level SAFe is for smaller value streams of around 100 people or fewer, while four-level SAFe is for large projects including hundreds or more practitioners per value stream.  

In a recent webinar, Leffingwell said that if you’re building systems with 50 to 100 people, or maybe multiple instances of 50 to 100 people who are working on a set of SKUs or products in a suite (and there’s not a lot of dependencies between them), then you’re good to go with three-level SAFe because you don’t have to coordinate much between the different products.

If you’re building a larger, more interconnected system, four-level SAFe will be more appropriate. In the webinar, Leffingwell used the example of a satellite system that does geophysical data acquisition. The system has three major components:

  • The satellite
  • The ground station
  • The web farm that delivers the geophysical data to a variety of consumers

These are large, complex, dependent systems, with hundreds or even thousands of people involved, and coordination between them requires additional mechanisms.

Should you upgrade your earlier version of SAFe to 4.0?

Although the most obvious change to the Big Picture is the addition of the new Value Stream layer, there are also many changes that apply to the more familiar three-layer model from earlier versions of SAFe. So should you upgrade?

“Absolutely,” says Peter Vollmer, solution architect and distinguished technologist at Hewlett-Packard Enterprise.

“SAFe is a framework that gives organizations the knowledge to apply the model in a way that fits their needs. You know your own situation best, so with an understanding of the SAFE principles, you should be able to make the right decisions. I recommend that you upgrade your knowledge to SAFE 4.0 so that you can benefit from the lessons learned from SAFe 3.0, and the merge with systems engineering.”

If you're moving to SAFe for the first time, read Anthony Crain's "4 biggest challenges in moving to Scaled Agile Framework (SAFe)."

What do you think of SAFe 4.0? Have you started updating your SAFe roles and processes to 4.0? We'd like to know; please add a comment below.

Topics: Agile