System change is not an excuse to avoid a difficult people decision; it is the discipline of locating the cause before making one.
Why it matters when the plan changes
When a plan slows, the two reflexes are to add process or to change the person. Both are high on the ladder and both are usually reached before anything below them has been tried. Where a plan runs through interfaces rather than through a single line of authority, cheaper structural moves such as resequencing or reassigning a dependency are usually available before any change of person is warranted. Sull, Homkes and Sull found only 9% of managers could rely on colleagues in other functions all the time, the condition the lower rungs address. The ladder exists to make that skipping visible, and to force the question of whether the cheapest credible move has been ruled out.
The tension runs the other way as well. A ladder that always recommends clarifying becomes a way of avoiding a decision that the evidence supports. A responsible recommendation keeps several rungs available and states why one should be tested first, including the case where the top rung is the honest answer.
In practice
A regional lead is slow to decide under a new operating model. The ladder asks first whether the priority is clear, then whether support exists, then whether the sequence is achievable, then whether the decision sits with the right person. Three of the four turn out to be untested. The move that looked like a leadership change was a reassignment.
Evidence
Where a plan runs through interfaces rather than through one line, the cheaper structural moves are usually available before any change of person is warranted.
Donald Sull, Rebecca Homkes and Charles Sull, Why Strategy Execution Unravels and What to Do About It, Harvard Business Review (2015)
What it cannot tell you
The ladder orders possible responses; it does not diagnose which rung fits a given case, that judgement depends on evidence about the actual cause of friction. It is silent on timing: testing lower rungs first can cost time a plan does not have, and the ladder gives no rule for when urgency should override sequence.
How Atlas reads it
Every recommendation names a rung and says why that one should be tested first. The ladder is applied to the arrangement around a person rather than to the person, which is what keeps a recommendation about the work. Changing leadership stays on the ladder because sometimes the demand and the evidence warrant it, and pretending otherwise would be its own kind of avoidance.
Questions
Clarify the priority or mandate, support with evidence or capacity, resequence the order or scope, reassign a decision or dependency, redesign a role or interface, and only then change leadership. Sull, Homkes and Sull's 2015 Harvard Business Review research on strategy execution supports testing the lower rungs first, since interface problems are usually cheaper to fix than people problems.
Because it is the most expensive and least reversible move, and because it is frequently reached before the cheaper ones have been tested. Last does not mean never. It means the case for it has to survive the questions that sit below it on the ladder.
Yes, when the evidence supports it. The ladder is an order of consideration rather than a mandatory sequence. What it forbids is skipping silently: a recommendation that starts at the fourth rung has to say why the first three were ruled out.
It applies to arrangements, usually a team, an interface or a role, rather than a person. Sull, Homkes and Sull found only 9% of managers could rely on colleagues in other functions all the time, which is why friction that presents as an individual problem often resolves through a redesigned forum or a reassigned dependency.
By what the reading says is causing the friction. Unclear priority is answered by clarifying, an unowned interface by reassigning, a role the plan has outgrown by redesigning. Matching the rung to the cause is the whole point, and mismatches are why interventions fail quietly.