I’ve sat through enough sprints to have an opinion about every Scrum meeting. Overall, I think Scrum does more good than harm. This post is the case against it anyway.
Not because I’ve changed my mind, but because Scrum’s costs are real and rarely counted. The benefits get the conference talks and the certificates. The costs get absorbed, one recurring calendar invite at a time.
What Scrum gets right
Scrum’s great strength is that it doesn’t let development run too far without a critical look. Every sprint ends with the work on the table and a question the team can’t dodge: is this still going in the right direction?
That’s also the guide’s own case for short sprints: more learning cycles, and less effort put at risk between them. Anyone who has watched a project vanish for six months and come back as the wrong thing knows what that’s worth. It’s the main reason I think Scrum does more good than harm.
The cost nobody budgets

Most practices fail passively. Neglect a team wiki, and it goes stale; neglect a style guide and people quietly stop following it. Scrum fails actively. Neglect it, and the meetings keep happening, on schedule, whether or not anyone has anything to say.
The numbers are bigger than most people assume. For a one-month sprint, the guide caps Sprint Planning at eight hours, the Sprint Review at four and the Sprint Retrospective at three. Add a fifteen-minute Daily Scrum on every working day, and you’re at about twenty hours out of roughly 160. That’s an eighth of the working month, per person, before anyone has refined a single backlog item.
Those are ceilings, not targets. For shorter sprints, the guide doesn’t even set a separate limit; it only says the events usually run shorter. But a ceiling nobody watches tends to become the target. Parkinson’s law says work expands to fill the time available, and meetings with a reserved slot are no exception.
So Scrum needs vigilance. Someone has to keep asking whether each of those hours is still earning its place. That vigilance is a cost too, and it never appears in the pitch.
The vigilance has a job title.
Scrum doesn’t leave that vigilance to chance. When the guide sums Scrum up in a nutshell, the first thing it says is that Scrum requires a Scrum Master. It makes that person accountable for the team’s effectiveness, and among the listed duties is making sure every event happens, stays productive and stays inside its timebox. That’s the vigilance this post is about, written up as a job description.
The same section also makes the Scrum Master accountable for establishing Scrum as the guide defines it, and for improving the team’s practices within the framework. In other words, a Scrum Master can change everything except Scrum.
That’s what makes the role different from a technical one. I’m certified myself, so I want to be fair here. When I recommend against a platform I’m certified in, I’m still doing my job; my certificate vouches for knowledge of a product that exists whether anyone hires me or not. A Scrum Master certificate vouches for a role the framework itself says is required. A Scrum Master who concludes that a team needs less Scrum has argued themselves out of the role.
This isn’t about the people. Many Scrum Masters are thoughtful and would gladly drop a ritual that isn’t working. It’s about the role, which, as written, has no room for that conclusion. The one person the guide holds accountable for the team’s effectiveness is the one person whose job rules out recommending less Scrum.
Scrum as defined

The guide is explicit about where it stands. It calls the framework immutable, and its closing note puts it plainly: “While implementing only parts of Scrum is possible, the result is not Scrum.”
That makes Scrum very hard to criticise. If a team runs Scrum and fails, the failure belongs to the team: they weren’t really doing Scrum. The framework itself can never be wrong. Anyone familiar with the No True Scotsman fallacy will recognise the shape. The vigilance requirement makes this worse, not better. If Scrum only works when someone watches closely enough, every failure can be blamed on someone not watching closely enough.
Scrum is also older than most people think. Ken Schwaber and Jeff Sutherland developed it in the early 1990s and first presented it at the OOPSLA conference in 1995, six years before the Agile Manifesto. The name goes back further, to a 1986 Harvard Business Review article by Hirotaka Takeuchi and Ikujiro Nonaka, “The New New Product Development Game”, which compared a newer way of developing products to a rugby team advancing together, rather than a relay race. Scrum is older than “agile”.
And it has barely changed. The official guide was first written in 2010 and has been revised five times since, most recently in 2020, in what its authors describe as small, functional changes. The 2020 revision made Scrum less prescriptive: it dropped the three Daily Scrum questions and cut the guide to under 13 pages. In 2025, Jeff Sutherland and two co-authors published a Scrum Guide Expansion Pack, but it’s a supplement; the 2020 guide is still the official definition.
Look at the pattern, though. Over the years, the revisions tightened how you run the events and then loosened it again, but none since 2011 has made one optional. For fifteen years, the framework that asks teams to inspect and adapt everything hasn’t once concluded that one of its own events could go.
The best case for the meetings
The strongest defence of Scrum’s events is in the guide itself: they exist to create regularity and keep other meetings to a minimum. That’s a fair point, and I’ll concede it. Teams that abandon scheduled meetings rarely stop meeting. Coordination moves into ad hoc calls, chat threads, and interruptions that are harder to see and that nobody counts.
But that’s an Argument for counting the cost of meetings, not for exempting them from the count. A meeting that replaces three others has earned its slot. A meeting held because the framework says so, whether or not there’s anything to discuss, hasn’t.
Scrum’s best idea, applied to Scrum.
I still think Scrum does more good than harm. Its best idea is simple: don’t let the work run too long without a critical look. My complaint is that it applies that idea to everything except itself. The events, the roles and the framework stay fixed, and the only thing allowed to fail is the team.
So who inspects Scrum? By the guide’s own rules, nobody: the retrospective can change how a team does Scrum, but the framework itself is immutable. In practice, the team has to do it, and that takes the vigilance this whole post has been about. Needing it isn’t a sign you’re doing Scrum wrong. It’s part of the price, and it’s worth paying knowingly: give Scrum the same critical look it gives everything else.
Written with AI assistance.
Comments