Why many people don’t like SAFe, and why I do

Wowie, am I gonna catch it now!

I’ve had some smart people ask why I’m anchored on the Scaled Agile Framework (SAFe), so I thought I’d jot some things down and begin to set out my position.

I’m 99.99% + confident that there are folks out there who can make strong arguments as to why I’m wrong on each of these points, and I hope to hear from you and learn something.

Why I think people don’t like it (feel free to add your reasons in the comments below):

Good reasons:

  • In spite of its origins as a ‘toolbox’ that can be applied by skilled professionals as needed and useful on a case-by-case basis, we’re seeing too many prescriptive ‘vanilla SAFe’ implementations without building any effective mechanism to learn and adapt as the implementation progresses.

  • SAFe reifies a one-way flow of ideas to work, leaving the people at the team level as a simple execution engine for the ideas developed by people further up the hierarchy for them. (Yes, I know about I & A and ‘innovation riptides’ and how they usually get implemented.)

  • SAFe relies on broadly-stated commitments to shift culture and the role of management, without any actual and proven learning path to change behavior – which risks leaving command-and-control in place atop an process framework designed for collaborative work.

  • Following on the above, SAFe is explicitly hierarchical, and saves more-important decisions for people further up the hierarchy. (I also say this below as a less-good reason).

  • SAFe doesn’t really dig into what makes teams at any level function well, other than (again) some bullet points. There’s a whole body of work on high-performance teams which I don’t (yet) see taken adequately into account.

Less-good reasons:

  • Following on the above, SAFe is explicitly hierarchical, and saves more-important decisions for people further up the hierarchy. (Didn’t you say that above? Why is this both a good reason and a less-good one? Because while our institutions are overly hierarchical, they still require hierarchy and role differentiation to scale. Anarcho-syndicalism only works in Monty Python movies)

  • SAFe isn’t LESS or DAD or Scrum at Scale or “my version of a scaling framework” (see: my discussion of how agile coaching is like 17th century martial arts in Japan). That doesn’t mean that any or all of those frameworks aren’t valuable, or in the right context, possibly more useful than SAFe. But the endless framework/approach wars are tiresome to those of us who practice agile – imagine what they’re like to outsiders.

Why I like it (again, jump in here, please).

When you boil it down to its core concepts and render all the fat and meat off the bones (sorry, my vegan friends), what’s left is pretty universally useful.

  • We want an n-level hierarchy of work which is represented by nested kanbans.

  • We want to work on a common cadence so we can share and transfer work more easily.

  • We want to collectively plan at each level as frequently as makes sense.

  • We want to use lean-agile principles (avoid delay, avoid waste, work sustainably, prioritize valuable work, manage batch size at each level to maximize flow) as much as we can.

  • We want to align along the path to value (this is not as easy as many people think).

  • We want top get the organization thinking about value at every level.

  • It’s popular. Why is popular good?

    • Because I can hire people who are likely to have worked in a reasonably similar environment

    • Because I can hire SME’s to help implement and manage it and the pool of knowledgeable SME’s (I didn’t say competent ones…) is large

    • Because other people have done it first, made the easy mistakes and I can learn from their experience and make different mistakes. Don’t underrate this – it’s super important.

  • It’s evolving into a catch-all for the best (or more popular…) practices in agility. As SAI is evolving SAFe, they are taking steps to pull in from concepts that are showing legs in the real world – product thinking, DevOps (as broadly defined), etc.

What don’t I like about it?

  • The prevalence of ‘bolting SAFe on’ that I see and hear about.

  • The emphasis on Big Upfront Transformation, rather than stepwise, intrinsic change.

  • The overemphasis on bringing in The Pros From Dover and the lack of emphasis and clarity on developing internal change agents.

  • The lack of a real path to mindset change and ultimately cultural change (note that no one else has them either).

  • The emphasis on ‘value” in environments where value is impossible to measure or maybe even define meaningfully – I prefer Yakyma’s emphasis on ‘outcomes’ here.

All in all, I’m happy to stay ‘SAFe-ish” and be part of the SAFe community, the members of which are constantly sharpening my thinking. I’ve found it a clarifying and useful tool in thinking about how to move organizations toward agility, and I find it something that real-world customers ‘get’ quicky, and so it offers a path off the X – a quick first step – for people working in traditional/normal hierarchical organizations. And as we all know, that first step is far and away the hardest.

Of course, I’ll constantly be using it as a framework I’ll pull to where my clients are – rather than the other way around. And I hope to keep learning SAFe, discovering problems with the framework, and proposing solutions for a long time.