Most organisations can name who produces the evidence. Far fewer can name who is obliged to act on it.

Ask an organisation who owns impact and you will usually get a name quickly. There is a Head of Impact, or a MEL lead, or a director with impact in their title.

Then ask a second question: what can that person decide?

The room goes quieter. Because in most organisations the honest answer is that they can decide what gets measured, how it gets measured, and what the report says. They cannot decide whether the programme continues, how the budget moves, who is hired, or whether a partnership ends. They own the evidence. They do not own any decision the evidence bears on.

This is the mandate–responsibility mismatch, and it is the most common structural flaw we find. It is not a failure of the individual. It is a design fault, and it produces predictable symptoms.

What the mismatch produces

Escalating rigour with flat influence. A capable impact lead whose findings keep not landing will usually respond by improving the findings — better methods, tighter analysis, more compelling presentation. None of it helps, because the constraint was never credibility. Effort rises, influence does not, and eventually the person leaves.

Impact as a veto nobody has. Ask what happens if impact evidence says a programme is not working, and listen for whether anyone can actually stop it. Usually the finding becomes an input to a conversation held by people with other priorities and firmer mandates.

The parallel committee. Organisations often respond by creating an impact committee. This frequently makes things worse: it moves impact into a forum where it can be discussed thoroughly and decided nowhere, and it relieves the main governance body of any obligation to look at it. A separate forum for impact is usually evidence that impact has been separated from the organisation.

Governance is not a committee

The word "governance" invites people to think about structures — who sits on what. The useful definition is narrower and more awkward: governance is the allocation of decision rights. Who may decide what, on whose evidence, at what point, and what happens if they decide badly.

Under that definition, three questions matter more than any org chart:

  • Which body decides? And critically, is it the same body that reviews financial and operational performance? If impact is reviewed somewhere else, it will lose every time the two conflict — not because anyone chose that, but because the forum with budget authority is the forum that decides.
  • What is it empowered to do? Note, advise, approve, halt, reallocate. These are wildly different, and organisations are routinely vague here in a way they would never tolerate on financial delegations.
  • What must be in front of it? If a decision can legitimately proceed without impact evidence, it will proceed without impact evidence whenever the timeline is tight. Which is always.

The test that exposes it

Take a real decision made in the last quarter — a genuine one with money attached. Now trace backwards. Was impact evidence in the room? Was it required to be? Who would have been accountable if it had been ignored?

In most organisations that trace terminates quickly. The evidence existed. It was available. It was not required, nobody was accountable for consulting it, and the decision was sound by every other measure.

Nothing went wrong. That is the point. The system performed exactly as designed, and the design has no place for impact in it.

Fixing it moves power

The remedy is not complicated to describe. Put impact review on the agenda of the body that already holds budget authority. Specify what that body may do. Make a defined class of decision unable to proceed without impact evidence. Give a named role accountability for the decision rather than the report.

It is not complicated. It is difficult, because every one of those moves takes discretion away from somebody who currently has it. This is why measurement is easy to fund and governance is hard: a new dashboard costs nobody any authority, and a decision right always comes from somewhere.

Which is also why it works. Organisations that make this change stop having the conversation about why their impact data does not land, because the structural reason it did not land has been removed.