Most organisations measure how long it takes to deliver things. Very few measure how long it takes to decide them. Yet decision latency – the time between a decision becoming necessary and it being made – is one of the largest hidden costs in any operating model.

Every week a decision waits, teams stall, options narrow, costs accumulate and good people get frustrated. In transformation programmes the effect compounds: one delayed decision can hold up several workstreams at once.

This report looks at where decisions get stuck, and at how clear decision rights and a few deliberate design choices let an organisation decide in days rather than quarters.

The hidden cost of slow decisions

Decision latency rarely shows up in a report. It appears instead as symptoms: work that is “on hold pending approval”, steering groups whose agendas roll over month to month, teams duplicating effort because nobody settled which option to pursue, and talented people leaving because they’re tired of waiting.

It also changes behaviour. When decisions are slow, people stop raising them – they work around the problem, or make the decision informally and hope nobody asks. Neither is good governance.

“When decisions are slow, people stop asking for them. That’s worse.”

Where decisions get stuck

In our experience, slow decisions almost always come from one of 5 causes.

1. Nobody is clear who decides. The decision has several stakeholders and no owner. Each assumes someone else will make the call, or that it needs everyone’s agreement.

2. Escalation is the default. Decisions that could be made at team level travel up the hierarchy because nobody is confident they have the authority. Senior forums fill with decisions they shouldn’t be taking.

3. Forums discuss but don’t decide. Meetings exist to review, update and align. They rarely end with a recorded decision. Items roll over from one month to the next.

4. The evidence arrives late. The forum meets, but the information needed to decide isn’t ready, so the decision is deferred to the next cycle – often a month or a quarter away.

5. Nothing says when to stop. Without agreed criteria for changing course, decisions to stop or redirect work are avoided rather than made.

6 design choices that speed decisions up

1. Map decision rights explicitly

For the decisions that matter most to your operating model, write down who decides, who must be consulted, and who needs to be told. Keep the list short – the 20 decisions that most affect pace are more useful than a comprehensive matrix nobody reads.

Frameworks such as RACI can help, but the critical part is the single name in the “decides” column. If a decision has more than one, it has none.

2. Set thresholds, not approvals

Instead of requiring approval for whole categories of decision, set thresholds: below this value, this risk or this impact, the team decides; above it, it goes up. Thresholds push decisions to where the information is and keep senior forums for the decisions that genuinely need them.

3. Make every forum a decision forum

Every recurring meeting should have a clear statement of the decisions it exists to make. If a forum doesn’t produce distinct decisions, merge it with another or stop it. We apply this test to our own governance: forums that don’t decide are a cost, not a control.

4. Keep a decision log

Record every significant decision: what was decided, by whom, on what evidence and when. It sounds administrative. It’s one of the most powerful tools for pace, because it stops the same decision being reopened and makes it obvious when decisions are waiting.

5. Prepare evidence before the meeting, not for it

Many decisions are delayed simply because the analysis isn’t ready. AI is now genuinely useful here: synthesising data, drafting options and summarising implications, so the people deciding spend their time on judgement rather than assembling the facts. The aim is that every decision arrives with the evidence already in place.

6. Agree stop criteria up front

For every significant initiative, agree in advance what evidence would justify changing course or stopping. It makes the hardest decisions easier, because the argument has already been had.

Measuring decision latency

What gets measured gets attention. Try tracking 3 simple measures for a quarter:

Measure	What it tells you
Days from decision raised to decision made	How fast your operating model actually decides
Proportion of decisions deferred at least once	How often forums meet without the evidence they need
Proportion of decisions escalated above their threshold	Whether authority is clear and trusted

Most organisations that do this are surprised by the results. [EVIDENCE NEEDED – optional: add anonymised client data on decision latency if available]

Self-assessment

Ask your leadership team to answer these honestly:

1. Could we name who owns each of our 10 most important recurring decisions?

2. Do our senior forums mostly make decisions, or mostly receive updates?

3. When a decision is deferred, do we know why?

4. Do teams know what they’re allowed to decide without asking?

5. Do we have agreed criteria for stopping our major initiatives?

If more than 2 answers are no, decision latency is probably costing you more than you realise.

Pace is designed, not demanded

Telling people to decide faster rarely works. Designing an operating model where it’s clear who decides, where evidence is ready on time and where forums exist to decide does. That’s where pace comes from – and in a period of rapid change, it’s one of the most valuable things an operating model can provide.

Common questions

What are decision rights in an operating model?

The explicit allocation of who makes which decisions, who is consulted and who is informed, usually with thresholds that determine when a decision moves to a higher level.

How do you speed up decision-making in an organisation?

Name a single decider for key decisions, set thresholds so routine decisions stay with teams, make every forum responsible for specific decisions, keep a decision log, prepare evidence in advance and agree stop criteria up front.

What is decision latency?

The time between a decision becoming necessary and it being made. It’s rarely measured but drives delay, cost and frustration across operating models and programmes.

Is RACI enough to fix slow decisions?

RACI helps clarify roles, but on its own it often becomes a document nobody uses. It works best when limited to the most important decisions, with a single named decider and clear thresholds.

What this looks like in practice

For a marine insurer going through a merger, we helped replace long waterfall approval cycles with agile delivery embedded as business as usual – getting live reporting to business units that had waited years, and tracking 74 concurrent initiatives through to delivery. Read the case study →