Interview: Value Stream Thinking – A New Model for Product Development
Peter Vollmer (Agile Point of View), Zach Brown (Atlassian), and Renaud Granier (Accenture) met at the SAFe Summit in Amsterdam for an in-depth conversation about Value Stream Management and Value Stream Thinking.
What is Value Stream Thinking?
While Value Stream Management focuses on tooling and automation, Value Stream Thinking is about applying a new model to product development. Peter Vollmer explains that traditional lean manufacturing concepts don’t directly translate to software product development — a highly complex, iterative process driven by learning cycles. Value Stream Thinking bridges that gap.
Peter illustrates this with a powerful historical analogy: during the 1854 cholera outbreak in London, most people believed bad air was the cause of the disease. Researcher John Snow took a different approach — he simply mapped the deaths on a street map of the Soho area. The visualization immediately revealed a cluster around a single water pump on Broad Street. When the pump was shut down, the outbreak stopped. Snow didn’t have better data than anyone else — he had a better model. That’s exactly what Value Stream Thinking offers: a model that makes the invisible visible, so you can find and fix the real root causes faster.
Four Perspectives on Value Streams
The model introduces four key perspectives:
- Enterprise perspective — the end-to-end view across domains
- Domain perspective — how individual teams and systems contribute
- Performance & observability metrics — using flow and DORA metrics to diagnose bottlenecks
- Way of working — identifying where agile, SAFe, or project-based approaches are used and where transformation makes sense
The Role of Tooling
Both Peter and Zach agree: no single tool does this well yet. Today, teams rely on Mural, Conceptboard, and similar visualization tools. However, Atlassian is actively working toward connecting process mapping tools with live data, AI-driven insights, and consistent reporting templates to support faster learning cycles.
The goal is not to have a tool that provides all the answers, but one that surfaces the right data — so that, like John Snow, practitioners can identify root causes faster and act on them sooner.
Watch the Full Interview here: