From Program Kanban to Portfolio Flow
I revisited my 2011 talk on scaling Kanban from team to program and portfolio. What still holds up, and what I would change today.
I recently revisited an old presentation I gave in 2011 about using Kanban to manage product development flow at the program and portfolio levels.
The context at the time was enterprises asking a very familiar question:
“Doing agile at the team level is fine. But how do we deal with the things we cannot get into the team?”
Fifteen years later, this is still the question.
The labels changed. Today we talk about product operating models, AI transformation, outcome orientation, and operating systems for delivery. But the core challenge is still flow across teams, shared resources, and strategic bets.
What Aged Well from the 2011 Talk
1. Limit work size at all levels
One of the first recommendations in that deck was to identify minimally valuable features and limit the size of work at every level.
That is still foundational.
If your portfolio items are huge, everything else becomes theater:
- forecasts are fiction
- dependencies explode
- feedback arrives too late to matter
The most practical move is still to split work into independently valuable slices and manage those slices as first-class flow items.
2. Visualize flow, not just status
The deck put heavy emphasis on visualization and cumulative flow, including a program-level aggregate CFD.
I still believe this is one of the fastest ways to create executive clarity.
Without flow visualization, leadership meetings default to stories and opinions. With flow visualization, you can discuss:
- where work is aging
- where WIP is piling up
- where throughput is dropping
This is exactly why I keep pushing teams toward active flow management at the portfolio level, not passive dashboards: Actively Managing Portfolio Flow
3. “Agile teams + waterfall release” is a smell
Another slide called out a smell I still see all the time: synchronized handoffs between teams that create a waterfall release pattern.
Many organizations still run local agility and global waterfall.
They improved team mechanics, but they did not redesign the end-to-end system. This gap is where most transformation effort stalls.
4. Feature teams and cross-team task forces beat component silos
The 2011 storyline moved from component coordination toward feature flow and cross-product collaboration.
That is still the right direction.
When organizations tell me “we do not have the political power to create true end-to-end feature teams,” I still recommend pragmatic stepping stones:
- temporary cross-product task forces
- explicit limits on concurrent cross-team initiatives
- deliberate dependency reduction work
Not perfect, but materially better than pretending coordination overhead is free.
5. Classes of service for shared resources are not optional
The presentation included risk profiling and classes of service for downstream shared resources.
This remains underused and high leverage.
If everything is “top priority,” nothing is.
Classes of service force explicit tradeoffs, improve predictability for fixed-date commitments, and reduce the hidden queueing that kills strategic initiatives.
What I Would Add Today
If I were giving that talk now, I would add five updates.
1. Connect program flow to portfolio flow by design
In 2011, many organizations were still figuring out program-level flow. Today, we need an explicit bridge from team and ART flow to investment flow.
That is where portfolio Kanban, flow metrics, and governance policies must connect: Improving Portfolio Flow Using Flow Metrics
2. Measure flow with intent, not as vanity metrics
I would make flow metrics operational and decision-oriented:
- WIP for focus and exposure control
- cycle time for service expectations
- throughput for planning and capacity shaping
- work item age for early warning
I covered this in both Scrum and SAFe contexts: Improving your SAFe Implementation with some additional Flow metrics
3. Pair flow with outcome steering
Flow alone is not enough. A faster feature factory is still a feature factory.
This is why I now emphasize linking flow to outcome hypotheses, strategic intent, and operating model choices, not just delivery mechanics.
4. Use a real management cadence
The portfolio does not improve because you have a board. It improves because leadership uses a disciplined cadence to make tradeoff decisions.
For most organizations, that means:
- weekly operational portfolio review (right-to-left, finish-first)
- monthly strategic review (investment direction, constraints, policy changes)
5. Treat AI-era work as flow design, not tooling rollout
Today, many leaders assume AI value is mostly a tooling problem.
My experience is the opposite: AI impact is constrained primarily by flow and operating model design.
If your organization cannot manage WIP, dependencies, and decision latency, adding AI mostly accelerates local noise.
Related read: AI Isnt Failing. Our Operating Systems Are.
A Practical 30-Day Move
If this resonates, here is a practical 30-day experiment:
- Map your current portfolio and program flow end-to-end.
- Add explicit WIP limits at portfolio and program levels.
- Introduce classes of service for fixed-date and high-risk work.
- Run a weekly right-to-left flow review with decision logging.
- Track aging and throughput trend lines for one month.
At the end of those 30 days, you will not have “solved scale.” But you will have moved from narrative to evidence, and from theater to traction.
Flow Discipline Still Wins
That 2011 deck ended with “guidance for real enterprises” and “pragmatic approach to change.”
I still stand behind that.
Frameworks come and go. Terminology evolves. But if you want better strategic outcomes, you still need to manage end-to-end flow deliberately and continuously.
If you want to go deeper on your current state, start here: Assessing Your Product Portfolio Agility
Improving portfolio flow requires both the right practices and the right mindset. Explore Portfolio Agility advisory or let’s talk.
A free email course on building an agile portfolio operating model — from recognizing your portfolio to achieving true strategic flow.
Yuval Yeret helps product and tech leaders move from agile theater to evidence-informed delivery. Work with Yuval →
