Counting the Days: Decision Latency as a Hidden Cost in UK Service Relationships
Photo: business meeting calendar deadline clock approval documents, via img.pikbest.com
Most organisations can tell you what they pay a service provider to the nearest thousand pounds. Very few can tell you how long it takes them to approve anything. That asymmetry is worth examining, because in many UK service relationships the cost of a slow decision exceeds the value of the concession that was negotiated to secure it.
Two clocks, not one
Decision latency is not a single measurement. It is at least three.
Queue time is the interval between a request arriving and anyone giving it attention. It is a function of diary pressure, unclear ownership and the quiet hope that the issue will resolve itself.
Deliberation time is the interval in which the decision is genuinely being considered. This is the only one of the three that can add value.
Communication lag is the interval between a decision being taken and the people who need it being told. It is almost never deliberate and almost always invisible.
Organisations that measure latency usually measure the whole span and conclude that decisions take too long, without establishing which of the three is responsible. The remedies are entirely different. Queue time responds to ownership and scheduling. Deliberation responds to better information. Communication lag responds to a habit of announcing outcomes the same day.
Where the cost actually sits
The costs of delay rarely appear on an invoice, which is why they survive scrutiny. They appear instead as teams held in a state of readiness, mobilisation dates that pass unused, option windows that close, duplicated preparatory work and informal workarounds that later have to be unpicked.
Consider an illustrative example. A change is priced at a modest premium for early mobilisation, and that premium is negotiated down over three weeks. During those three weeks the provider's delivery team remains on standby, the client's programme waits, and a second workstream is rescheduled. The saving is real and visible; the cost is real and diffuse. Boards approve the first and never see the second.
There is a second cost that is harder to quantify and easier to feel. Providers learn how long their clients take. Some build the delay into their plans and their prices. Others stop proposing improvements, because the effort of securing approval outweighs the benefit of the change.
Measuring without building a bureaucracy
The temptation is to impose a decision-tracking system across the whole organisation. Resist it. A sample is sufficient.
For a quarter, log five timestamps against a representative set of decisions: when the request was raised, when the information was complete, when the decision was taken, when it was communicated, and when it was acknowledged. Categorise each decision by type — commercial, technical, compliance, staffing. Then report the median and the upper decile rather than the average, which is distorted by a handful of easy cases.
The output is rarely a number that surprises the board. The value lies in the segmentation. In most organisations, one category of decision accounts for the majority of the delay, and it is usually not the category anyone expected.
Deliberation or drift?
Not every delay is a failure. Some decisions deserve time, and a board that prides itself on speed is not necessarily well governed. Four questions help distinguish the two.
Is the decision reversible? Reversible decisions rarely merit extended deliberation; the cost of reversing is bounded.
Is the missing information genuinely material? If the decision would be the same with or without the information being sought, the delay is procurement of comfort, not of evidence.
Has anyone asked for the decision in the last week? If not, the matter has left the decision queue and entered the backlog, which is a different problem with a different owner.
Would a decision taken today be materially worse than one taken in a month? If the answer is no, the delay has become a cost with no corresponding benefit.
Setting expectations on the client side
Responsiveness is a client obligation as much as a provider one, and it is rarely written down. Three measures help.
Name an approver and a deputy for each category of decision, with the deputy empowered rather than merely informed.
Align delegated authority thresholds to the scheme of delegation already in place, and check that the thresholds reflect the size of decisions now being taken rather than the size of the organisation five years ago.
Publish an internal standard. A commitment to respond to a complete request within a stated number of working days costs nothing and changes behaviour, provided the word complete is defined and the standard is reported on.
Keeping judgement out of the bottleneck
Speed should not be achieved by centralising everything into a single approving individual. That trades one bottleneck for another and removes the deliberation that occasionally earns its keep.
The better pattern is pre-agreed parameters: standing approvals for recurring decisions within defined bands, an escalation clock that lifts unresolved matters automatically, and a decision log that makes the pattern of delay visible to the board without naming individuals. Over time the log becomes the most useful management information in the relationship, because it shows where authority, information and urgency have quietly diverged.
Measured quietly and reported honestly, decision latency is one of the few service metrics a client fully controls. That makes it a sensible place to start.