The Price of the Pause: Auditing Decision Latency in UK Operations

Why delay deserves a ledger

Most UK organisations can quote the cost of a hire, a supplier or a software licence. Few can state, with confidence, what it costs to wait for a signature. Decision latency is rarely recorded as a cost because it arrives in fragments: a finance director chasing a purchase order, a client wondering why a proposal has not arrived, a project team idling while a variation is reviewed. Individually these moments look like administrative nuisance. In aggregate they form a material drag on performance.

The purpose of an approval audit is not to celebrate speed for its own sake. It is to establish where delay is buying genuine control and where it is merely consuming time. For UK organisations operating in competitive markets, with tight client expectations and finite management capacity, that distinction matters. A slow approval route can be a sensible safeguard against risk. It can also be a habit that nobody has examined for years.

Map the chain before measuring it

Begin with decision classes rather than departments. Select three or four recurring categories: routine purchases above a defined threshold, client commitments such as proposals and statements of work, operational changes, and exceptions to standard policy. For each class, reconstruct the route that decisions actually take. Do not rely solely on the policy manual. Policy describes intent; the audit should describe practice.

For a sample of recent decisions, record the originator, every reviewer and approver, the sequence of handoffs, the systems used and the points at which the decision waited. Include informal gatekeepers: the executive assistant who schedules a director's review, the finance business partner who pre-checks a business case, the legal adviser whose comments arrive by email rather than workflow. These roles may not appear in a delegated authority matrix, but they shape elapsed time.

Pay particular attention to sequential steps that could run in parallel, committee dates that create artificial cut-offs, and thresholds that trigger a disproportionate level of scrutiny. A purchase of modest value that requires three signatures may be appropriate if fraud risk is high; it may be unnecessary if the same control is already provided by budget ownership and spot checks. The map should make those questions visible.

Quantify in working days, not impressions

Approval delay is often described in vague terms: slow, cumbersome, heavy. An audit needs a unit of measurement. Use working days, excluding weekends and public holidays, and define the clock precisely. Start when a decision is ready for approval, not when someone first drafts an email. Stop when the final authorised person approves, not when the originator receives an informal verbal yes.

For each decision class, calculate the median elapsed time, the upper quartile and the longest cases. Where possible, separate active time from waiting time. A decision that takes ten working days but only two hours of actual review is not a capacity problem; it is a queueing problem. Compare actual performance with any stated service standard, such as five working days for purchases below a threshold or two working days for client contract exceptions.

Link the numbers to consequence. For client commitments, count proposals that missed a client's decision window, start dates that slipped, or discounts offered to recover goodwill after a delayed response. For purchases, identify delayed operational improvements, expediting costs and repeat chasing by managers. Estimate opportunity cost with conservative assumptions and present a range rather than a single figure. The aim is not false precision; it is a credible order of magnitude that justifies attention.

Separate control from friction

Not every signature is a control. A useful test is to ask whether the approver has both the authority and the information to refuse or change the decision. If the answer is no, the step may be assurance, coordination or simple habit. Another test is whether the approval changes outcomes. A step that almost always approves, rarely queries and duplicates a risk already managed elsewhere is a candidate for redesign.

Classify each step in the chain. A genuine control reduces a defined risk, such as fraud, safety, regulatory breach or contractual exposure. An assurance step informs a senior person but does not alter the decision. A coordination step sequences work across teams. A friction step exists because it has always existed. The labels should be applied with evidence: approval rates, query rates, documented rationale, regulatory requirements and the judgement of the people involved.

This is where an approval audit differs from a simplistic delegation exercise. Removing a control without understanding the risk it addresses can be careless. Retaining a control that no longer earns its place is also a choice, with its own cost. The audit should inform a deliberate redesign, not a race to the fewest signatures.

The client-facing cost

Client commitments deserve particular scrutiny because delay is visible outside the organisation. A proposal that sits with legal for a week, a statement of work awaiting a commercial director, a pricing exception caught in a committee cycle: each may be internally rational and externally damaging. Clients do not see the approval chain. They see responsiveness.

Track the interval between a client-ready document and an authorised response. Compare it with the client's own timetable. In competitive tenders, a day can determine whether a submission is considered at all. In ongoing relationships, repeated small delays can erode confidence long before a contract is lost. The audit should therefore include lost and stalled opportunities, not only completed approvals. That wider view often provides the strongest argument for change.

A six-week audit routine

A practical audit can be completed in six weeks without becoming a major programme. In the first week, agree scope, decision classes and data sources. In weeks two and three, collect recent cases and reconstruct the chains. In week four, calculate latency, classify steps and identify the largest sources of waiting. In week five, hold short workshops with approvers and originators to test the findings. In week six, present a baseline, a control-and-friction assessment and two or three redesign options.

Useful measures include median cycle time, the proportion of decisions completed within target, the number of handoffs, the ratio of active to waiting time, rework rates and escalation frequency. Pilot any change on one decision class before applying it more widely. Adjust thresholds, parallelise reviews, pre-approve low-risk categories and publish clear service expectations. Assign an owner and review the measures quarterly so that the old habits do not quietly return.

The discipline is not to remove control but to place it where it earns its keep. An organisation that understands its approval chains can protect the risks that matter, release the ones that do not, and give its people the confidence to move at the pace its clients expect.