Resume.org 'Great Turnover' survey (2026)

A woman I will call a Scrum Master, because that is the title her company handed her, paid for a two-day course, passed a multiple-choice exam, and walked into a department she was now expected to transform. She had no team of her own and no standing with the directors whose behavior she was there to change. What she had was a certificate and a box on the org chart. Within a month the senior manager who controlled her review told her to keep the stand-ups and drop the part where the team decided its own work, because he wanted the old reporting back. She did what almost anyone with a mortgage does. She complied. Then she took it from both directions, from the manager who found her useful only when she obeyed, and from the practitioners online who looked at her team and pronounced that this was not real Agile.

That last verdict is the one I want to sit with, because I have delivered it myself, and it took me too long to see what was wrong with it. Last week I wrote about the executive who tries Agile, fails, and decides the method is broken, when what actually broke was his unwillingness to give up control. The practitioner who sneers at that executive is not standing on the higher ground he imagines. He is running the same play from the other end of the field.

Here is the first thing the purist gets wrong, and it is almost funny once you see it. Agile, whatever else it is, is supposed to be empirical and incremental. You look at where the work actually is, you change one thing, you check whether it helped, and you adjust from there. The whole philosophy is built on meeting reality where it stands and moving forward in small, tested steps. And yet the purist shows up to an organization with the least Agile posture imaginable, a demand that you conform to the ideal now and a verdict that anything less is failure. The incremental move his own philosophy rests on, the one that says you are here today and the next honest step is just there, is the one he never offers. He wants the finished state installed by Friday, which is the executive's quick-win fantasy in a different costume.

The second mistake follows from the first. Once you grade an organization on how faithfully it runs the ceremonies, you quit asking the only question that matters, whether the business actually got better. A team can run textbook Scrum and ship nothing a customer wanted. A team can break every rule in the framework and still change how it serves the people it exists for. The purist gives the first team top marks and the second a failing grade, and in doing so admits he was never measuring results at all. He was measuring obedience to a process. That is the same yardstick the controlling executive uses, simply held up to a different altar.

Underneath all of it sits a word I have to handle with care, because it carries more than it looks. Authority. In these roles, real authority has little to do with the box on the org chart or the title on the certificate. It is influence earned through expertise, the standing that makes a senior leader actually stop and listen to you. You cannot grant that in a weekend, and you certainly cannot sell it. The certification industry sells it anyway. It tells the market the role confers authority, that the holder of the credential ought to be heeded by dint of holding it.

And what does the credential require? Money, mostly. For most of these certifications there is no meaningful bar beyond the fee and a seat in the room. So the industry mints people who have a credential and no influence, drops them into organizations run by executives who guard every decision of consequence, and instructs them to move those executives. When they cannot, and of course they cannot, the same industry that sold the certificate clarifies that the failure was personal. They were not a true practitioner. The framework stays spotless. The invoices keep clearing. The blame rolls downhill to the cheapest and least powerful person in the building, and the product that actually failed never takes the stand.

What genuinely angers me is what this does to the people in those seats. We take someone with no authority, demand that they exert authority, and then hold them in contempt when they bend to the manager who signs their checks. They gave in because they wanted to keep a job and feed a family, inside a setup that loaded them with responsibility and withheld the power to carry it. To then call that person a sellout, or a fraud, is diabolical. They were arranged to fail, by the executive who would not loosen his grip and by the purist who would not accept anything short of perfection, and then handed the bill for both.

That is the seam between the two halves. The executive and the purist look like opposites. One calls Agile nonsense, the other treats it as scripture. They make the same mistake all the same. Each wants a transformation he can define from outside and impose whole, the executive by mandate and the purist by doctrine, and each refuses the slow, empirical, trust-built work of moving a real organization one honest step at a time. The change agent standing between them, the one actually at the point where the work gets done, is what gets crushed in the press.

If that reads like a quarrel from a passing era, look at what is assembling around AI right now. The certificates are already for sale. Prompt engineering credentials and AI transformation badges, with a new priesthood informing organizations that they ought to be AI-first and that anything less is a failure of nerve. The scolding keeps its old shape. You are not using it correctly. That is not real AI adoption. And the same trap is under construction for a fresh round of people who will be hired for a badge, given no authority, set down between executives chasing a quick win and purists guarding the doctrine, and blamed when the impossible declines to happen. We are running the Agile picture again on faster hardware, and so far we have taken nothing from the first screening.

So the honest version, for those of us who do this work, is harder and less flattering than a certificate. Meet the organization where it truly is, not where the framework insists it should be. Change one real thing and show that it helped before you reach for the next. Earn the influence rather than invoke the title, and accept that you will be judged, like everyone else, on whether anything improved rather than on whether the ritual was kept. And if you are going to certify a person and call them a leader, give them the standing to act like one, or drop the pretense that the certificate was ever the point. For the practitioners reading this, and I count myself among you, the last time you called something fake Agile, were you protecting the work, or protecting yourself from ever having to prove it produced a result?