Did You Know?
I recently discovered this article from SafeDelusion. It does a great job at highlighting the challenges of the Scaled Agile Framework (SAFe), suggesting that while it promises structured decision-making at scale, it often creates the illusion of empowerment rather than the reality.
Instead of speeding up organizational flow, SAFe can unintentionally (or in some implementations intentionally) reinforce bureaucracy, compliance-driven rituals, and centralized control.
This critique echoes what we teach in agile, which emphasizes that decision-making under uncertainty requires flow, decentralization, and fast feedback loops, not excessive governance.
Leaders must avoid confusing structure with true safety (agility). They might sleep soundly at night knowing all processes are in place, but in reality, the business is slowly dying a death.
So What?
Why does this matter for leaders?
- SAFe can slow down flow. While SAFe intends to align large organizations, its layers of approval and committees can delay decision-making at a time when speed is critical.
- AI widens the gap. With generative AI and machine learning compressing decision cycles, organizations bogged down by SAFe rituals will lag behind nimbler competitors.
- Employee engagement suffers. Kotter reminds us in Leading Change that urgency and empowerment fuel transformation. But if SAFe devolves into rigid governance, teams feel more like order-takers than innovators.
- Economic signals get buried. Reinertsen warns against proxy measures (like adherence to process) over true economic drivers like the cost of delay. SAFe’s ceremonies sometimes prioritize framework compliance over outcomes.
In short: SAFe may provide comfort to leaders, but it risks creating a false sense of agility.
Now What?
Instead of blindly adopting SAFe as the only model for scale, organizations should:
| Action | Why It Matters | How AI Can Help |
|---|---|---|
| Shift focus from framework compliance to flow | Improves adaptability and reduces delivery bottlenecks | AI analytics reveal true cycle times, queues, and cost of delay |
| Decentralize decision-making | Empowers those closest to customers and work | AI copilots deliver insights directly to team-level leaders |
| Measure economic outcomes | Shifts focus from “following SAFe” to delivering ROI | Predictive models quantify tradeoffs in time, cost, and value |
| Use fast feedback loops | Prevents bad decisions from compounding across the enterprise | AI-powered experimentation accelerates learning cycles |
Catalyst Leadership Questions
| Question | Probing Follow-up |
|---|---|
| Where are we following SAFe for compliance vs. customer value? | Can we point to a business outcome improved by this decision layer? |
| Which decisions could be safely decentralized today? | What information or tools would teams need to own them confidently? |
| Do we understand the cost of delay for our top initiatives? | If a program slips a quarter, what does that cost the business? |
| Are we optimizing for learning speed or framework compliance? | Where has SAFe slowed us down in the last year? |
| How could AI reduce our decision bottlenecks? | Can AI copilots surface the insights we’re waiting on governance for? |
SAFe is not inherently bad; in fact, Scrum gets the same scrutiny when we misapply the framework. Done well, SAFe brings order to complexity, but is that even possible?
When organizations mistake framework compliance for agility, they risk slowing down just as AI is speeding up the market. True business agility doesn’t come from rigid structures; it comes from flow, feedback, and empowered decision-making at every level.
Let's Do This!
SAFe is scaffolding; it can help you build. But if you keep it up forever, it starts blocking the windows.
In transformation work, “SAFe” decisions can be particularly risky because they lead leaders to believe progress is being made when, in fact, flow is slowing. In product development, avoiding risk entirely often creates bigger risks later.
The organizations that thrive will be those that embrace fast, economically informed, AI-assisted decisions at the edge of the organization.
Think of it like traffic, adding more stop signs might feel safer, but if the goal is to get people home faster, sometimes the best move is to build more innovative moving intersections that keep cars flowing or empower drivers to come up with their own rules.
Yes that is scary, but so is your business drying from calcified and ossified processes.