The Waiting Room Premium: How Slow Approvals Inflate UK Project Costs

The hidden line item

In UK commercial projects, poor decisions attract scrutiny. Slow decisions often do not. A delayed yes can be as costly as a wrong no. Every week an instruction waits for signature, preliminaries continue, site management stays mobilised, equipment hire rolls on and specialist suppliers hold capacity they could have released elsewhere. In 2026, with volatile material pricing and tight subcontractor availability, delay has a visible price even when it never appears as a single invoice line.

Under NEC4, early warnings and compensation events run to clocks. Under JCT, instructions and variations depend on promptness. Yet many clients still measure approval quality rather than approval speed. That is a commercial blind spot. A decision that is technically sound but made three weeks late can still damage margin, programme certainty and supplier goodwill.

What decision latency actually is

Decision latency is the elapsed working time between the point at which a decision could responsibly have been made and the point at which it was made. It is not the same as deliberation. Good governance needs time to test evidence, challenge risk and record reasoning. The problem is the time spent waiting for a meeting, a missing signature or a person who is not in the room.

It helps to separate three forms of delay:

• information latency: waiting for facts that were not supplied;
• authority latency: waiting for a named person or forum to decide;
• queue latency: waiting for the next scheduled meeting or approval slot.

In most commercial reviews, authority and queue latency dominate. The analysis has often been done. The decision is simply sitting in a tray.

A practical starting point is a decision log. For each material decision, record the request date, the date the pack became decision-ready, the decision owner, the authority level, the value and risk band, the date of decision, the outcome and any cost of delay. From that log, a client can calculate median latency and 90th percentile latency by category. The median tells you what normally happens. The 90th percentile tells you what happens when a project is under pressure.

The cost is not only direct

The direct costs are easiest to see: extended preliminaries, standing time, escalation claims and lost float. The indirect costs are often larger. Suppliers learn which clients decide slowly and price accordingly. They hold capacity, add risk allowances or deprioritise the slow-moving instruction when a better-programmed job arrives. Programme certainty falls. Contingency is consumed before the work starts.

There is also a governance cost. When decisions are slow, project teams create workarounds. Informal approvals creep in, records weaken and accountability blurs. The answer is not to abandon controls. It is to design controls that operate at the speed the project requires.

Delegated authority with guardrails

A delegated authority matrix should be built around value, risk and reversibility, not hierarchy alone. Low-value, repeat or reversible decisions can sit with a project manager or commercial lead within defined limits. Medium-value decisions can require a commercial and technical sign-off. High-value, high-risk or irreversible decisions can go to an executive or committee.

The matrix needs named deputies, clear quorum rules and standing delegations for defined periods. A two-key approval is often enough where organisations currently use five. Exception reporting, retrospective sampling and segregation of duties preserve control. Thresholds should be reviewed quarterly, not left untouched for years.

Governance is weakened not by speed itself but by unclear ownership. A decision made quickly by the right person, with the right evidence and a clear record, is stronger than a decision made slowly by a forum that cannot remember why it approved.

Shortening the cycle

Several practical measures shorten approval cycles without dismantling control. Agree a decision calendar at project start. Name decision owners in the contract administration plan. Use a standard decision request template so that packs arrive complete. Set a service-level expectation for response, such as five working days for standard decisions and 48 hours for time-critical ones. Allow asynchronous approval where the risk is low and the audit trail is sound. Use default escalation rather than default approval when a deadline passes.

For NEC4 projects, align the decision log with the early warning register. For JCT projects, align it with the instruction protocol. The mechanism matters less than the discipline of knowing where each decision sits and who must move it.

Measure what matters

A useful dashboard shows median latency, 90th percentile latency, the percentage of decisions made within the agreed service level, the number of aged decisions and the estimated cost of delay. Report it alongside quality measures such as rework caused by late decisions. Speed alone is not the goal. Predictable, well-recorded decisions are.

Conclusion

A slow yes is a commercial decision with a cost. Clients that treat it as one, measure it honestly and delegate authority with clear guardrails protect margin and supplier trust. The objective is not to decide faster for its own sake. It is to remove the waiting room premium that quietly accumulates while everyone assumes the decision is being handled.