Decision rights during a change: what you record and where

You record who may decide at what level, who must be consulted for advice, and which decision automatically escalates to the leadership team. That sounds simple, but it usually only becomes uncomfortable once the change becomes concrete and no one knows anymore who gets to make the call.

Why this is often not recorded

In a company of 50 to 300 employees, decision-making runs on gut feeling for a long time. The leadership team knows the people, the people know the leadership team, and most decisions get made in a hallway or a short meeting. That works, until the number of decisions increases and multiple layers of the organization are involved at the same time: a team leader who wants to test a new process, a department head who wants to shift budget, a director who wants to replace a supplier. Without recorded decision rights, one of two patterns emerges: everything comes back to the leadership team, or departments decide independently and contradict each other three months later.

The problem is often not unwillingness but language. Two managers use the same word for a different level of decision, and that is exactly why people mean different things by the same words without anyone noticing, until the decision has already been made.

What you record per layer

The target state works with four layers: vision, vision state, target, and target state. Decision rights do not belong in one document, but differ per layer.

For the vision and the vision state, the decision right almost always lies with the leadership team or owner. This is where direction is set, not execution, and that decision cannot be delegated without changing the vision itself.
For the target — the concrete result within a period — part of the decision right shifts to the level responsible for execution. Recording exactly what is delegated here, and what is not, prevents a department head from making a decision that actually affects the vision state.
For the target state — the roles, tasks, and way of working needed to get there — decision rights usually sit lowest in the organization, with whoever actually does the work. This concerns operational choices that the previous layers should not need to confirm.

The question asked per decision is not "who is responsible" but "which layer is affected if this decision goes wrong". That determines at what level it belongs.

An example from practice

A company with 120 employees decides to restructure customer service. The team leader is given the assignment to "reduce waiting time" and independently decides to hire two new employees and extend opening hours. Three months later it turns out that the leadership team actually had a different direction in mind: fewer people, more automation. The team leader's decision was not wrong — it simply did not fit a vision state that had never been explicitly shared at the level where the team leader operated.

Had it been recorded that staff expansion above a certain number of FTE must go back to the leadership team, and that the team leader is free to decide on opening hours and scheduling, this decision would not have been reversed three months later but would have been correct from the start. That is the difference between decision rights that exist on gut feeling and decision rights that are put on paper.

The double check and confidence gates as a control point

The eight dimensions and five confidence gates are not only a check on the content of the strategy, but also a natural moment to test decision rights. At every gate that is passed, a decision has been made: continue, adjust, or stop. Recording who may make that decision at which gate prevents a gate from being passed based on the enthusiasm of one person while the rest of the organization does not stand behind it.

This also touches the question of how you know a change is really working. Decision rights that are properly in place are visible in who can proceed without consultation and who cannot — and that is one of the early signals that a change is taking hold. Where everyone still looks upward for permission, that signal is not yet there.

What this prevents in the longer term

A strategy that does not translate into decision rights remains an intention on paper. That is precisely the mechanism by which a strategy remains a document instead of something people use daily: no one knows who may decide, so no one decides, and the document collects dust. Recording decision rights is therefore not a side issue next to the strategy, but one of the few things that determines whether it ever does anything.

Whether your vision is currently clear enough to actually assign decision rights to can be tested with six questions about the four layers in the free test "is your vision steerable".

What you can do now

Start with one recent decision that landed wrong or remained undecided for too long, and lay side by side who made it, who should have made it, and which layer — vision, vision state, target, or target state — was affected by it. That one example often already shows where decision rights still rest on gut feeling instead of on paper. The next step is the translation to capabilities: which roles and tasks the target state requires, alongside what is already in place today in terms of people and work. That difference, expressed in hours, is in the work scan on ftetoai.com.