SAFe is a named, off-the-shelf scaled-agile operating model that companies have adopted wholesale as a template for coordinating agile work across large organizations — referenced alongside a pun on the Safeway grocery chain, capturing its 'buy it off the shelf' character.
Sits alongside Operating Model as Starting Point, Not Rigid Blueprint's broader critique: SAFe, the Spotify Way, and the SVPG model are all cited as examples of companies importing a complete named framework as 'the right way to run product,' rather than treating it as a starting point to adapt. See Operating Model as Starting Point, Not Rigid Blueprint for the general pattern and its fallout.
Cagan cites SAFe as the paradigm case of the common, flawed response when a founder's personal approach to product stops scaling: imposing a heavyweight process. He invokes Steve Jobs' framing of this as "the disease of process people." The alternative Cagan argues for is not a bigger process but better leadership — specifically the coaching component of "Founder Mode" Misread as License to Micromanage — the Real Lesson Is Founder-Style Leadership — since process is what organizations reach for when they don't know how to scale coaching instead.
SAFe (Scaled Agile Framework) is, per Marty Cagan, not actually Agile at all — it's a process framework optimized for predictability rather than outcomes. Cagan calls its adherents "the disease of process people." The tension is captured by a sign Spotify hung in its halls: "100 percent predictability equals 0 percent innovation." Product companies that run on Product Model (vs. Roadmap Model) thinking optimize for outcomes and innovation; SAFe optimizes for schedule certainty, which makes it structurally opposed to the product operating model even when marketed as an "Agile scaling" solution.
The speaker is blunt that SAFe fails on its own terms as an Agile framework, not just as a product-model antithesis:
"Yeah, safe is not safe."
"It's only agile in marketing. Name only. There's no Agile there, at all. None of the principles."
This sharpens the existing critique: SAFe isn't merely incompatible with the product model, it doesn't deliver on the Agile principles it's named after either — making it a double failure (neither genuinely Agile nor product-led) rather than a legitimate alternative path.
Cagan reports that his firm treats a company self-describing as using SAFe as a reason to advise against taking them on as a client: companies still happy with SAFe have typically only encountered its marketing, not lived its consequences at scale. Satisfaction with SAFe is read as evidence the org hasn't yet hit the problems the framework causes, not as evidence the framework works.
Apply: treat enthusiastic self-identification with SAFe as an early-stage signal (dogma not yet tested by scale) rather than a sign of a mature or working process.
Cagan calls SAFe "literally waterfall," not agile, and argues it's adopted mainly by CIOs who prioritize predictability of delivery over innovation — the opposite trade-off the Product Model (vs. Roadmap Model) asks teams to make.
Cagan is unusually blunt about SAFe in this talk, calling it "absolute garbage" and "pure marketing," and rejecting the idea that it's meaningfully agile at all: "it's just waterfall, I mean it, it's literally waterfall." His diagnosis of why it spreads anyway: CIOs adopt it because it optimizes for predictability, not innovation — and per his framing, "100% predictability equals 0% innovation."
Из тем: The Product Operating Model