Incomplete Design Information

The situation

Paid for the risk, and kept it

Programme pressure had brought procurement forward. Structural design was resolved, the envelope broadly was, and the services design was not. Rather than delay the market approach, the decision was taken to tender on the information available and to develop the remainder during the contract.

That is a legitimate commercial decision. Nevertheless it is only legitimate when the client knows precisely what they are transferring and what it costs them.

Here they did not. Contractors priced the undefined services scope by loading it, then qualified their return so that if the developed design exceeded their assumption the difference came back as a variation. Therefore the client had paid a premium for the risk and had also retained it.

What we found

Incomplete is not the same as undefined

The distinction that mattered was between design that was incomplete and design that was undefined. Those are different problems and they carry different prices.

01

Substantial parts of the services design were incomplete but entirely predictable, which meant they could be scoped and measured with confidence rather than left to contractor assumption.

02

A smaller portion was genuinely undefined, mostly at the interface between landlord and tenant systems, where a commercial decision had not yet been taken.

03

Contractor qualifications did not distinguish between the two. Consequently the whole scope was priced at the risk level of its least certain part.

04

Provisional sums had been used as a substitute for scoping rather than as a controlled mechanism, with no conversion process attached.

What changed

Reduce uncertainty to the part that is uncertain

The client accepted a known exposure on one interface. Everything else stopped being a negotiation.

1

Measure what is predictable

The predictable scope was defined and measured, which removed it from the risk conversation entirely and from the price the contractors were loading.

2

Show the genuine unknown openly

Only the undefined element remained as a priced allowance, shown in the cost plan with a decision date against it rather than buried inside contingency.

3

State what transfers and what is retained

Tender documentation said plainly which risks were being passed to the contractor and which the client was keeping, so both parties priced the same thing.

4

Watch the pricing of uncertainty narrow

Contractor allowances for risk fell sharply, because the uncertainty they were being asked to price had been reduced to the part that was actually uncertain.

Why this happens

Two programmes, set by different parties

Design programmes and procurement programmes are set by different people, and they very rarely align. When they do not, the commercial function has one useful job: to state plainly what is not yet known, price it honestly and make the client's exposure explicit before they commit.

Cost planning and estimating carries that assessment through the design stages so the position is current when the market approach is made. In addition, tendering and procurement determines how the residual uncertainty is packaged, allocated and priced.

Questions

Asked about incomplete design

Can a budget be reliable when the design is incomplete?

It can be honest, which is more useful than reliable. An honest budget names each incomplete area, states the assumption used, attaches a value to it and shows what the total becomes if the assumption moves. That gives a client something to act on: which decision to take first, which survey to commission, which allowance to protect. A budget that presents an incomplete design as a single confident figure is not more reliable, it is simply less informative about where it might move.

What is the difference between a provisional sum and a contingency?

A provisional sum covers work that is known to be needed but is not yet defined well enough to price. Contingency covers work that may not be needed at all. Conflating them hides how much of your total is committed and how much is genuinely spare, and it is one of the most common weaknesses we see in cost plans. Keeping them separate, with named items behind each, is what lets you report movement meaningfully as the design resolves.

How should incomplete information be handled in the contract?

By naming it, valuing it and setting out the mechanism for resolving it, rather than by silence. Where information is missing and the contract does not acknowledge that, the risk falls wherever the drafting happens to leave it, which is rarely where either party intended. Defined provisional sums, clear information release dates and a stated basis for valuing the eventual instruction all reduce the argument later. None of that requires the design to be finished first.

Do contractors price missing information or ignore it?

Both, and which one you get depends on the market and the bidder. In a competitive market a contractor who prices the risk properly loses to one who does not, so the winning tender is frequently the one that assumed the missing information away. That price is not lower, it is deferred. Recognising which parts of a return depend on an unstated assumption is exactly what a tender analysis is for, and it changes how a recommendation should be written.

What should we resolve first when design is behind?

The items where a change in assumption moves the largest amount of money, which is not always the same as the items causing the most concern. We work through the cost plan and rank the open assumptions by value at risk, then advise which decisions or surveys unlock the most certainty per week of programme. That ordering is usually different from the design programme's own sequence, and it is where a commercial view earns its place.

Related scenarios