Taking over a development that is running over budget

A scheme in trouble arrives in a recognisable state. The spend is ahead of the programme, the contingency has gone, there is a list of variations nobody has valued, and the contractor and the client have different numbers for everything. Whether we come in as a replacement for the outgoing team or alongside it, the sequence is the same, and it is the sequence that makes delivery and cost control recoverable.

Nothing useful happens until the real position is established. Cutting cost from a position you cannot evidence produces an argument with the contractor and a second overrun. The order is: establish, separate, reset.

Establish the position before changing anything

Start with what is committed rather than what is spent. The contract sum, every instruction issued whether or not it has been valued, every variation claimed whether or not it has been accepted, the subcontract and supply commitments placed, and the fees and costs outside the building contract that are running regardless.

Then measure what is actually built. Not the percentage on the last application: the work in place, measured on site against the contract documents. On a scheme that has lost control this almost always differs from the certified position, in one direction or the other, and the difference is the first real piece of information.

Then set out what remains. What is left to build, what it will cost at today’s prices rather than the tender’s, what risks are still open and what the programme now says. The output is a forecast final cost that is defensible, which is the number everything else hangs off, as explained in forecast final cost.

Separate genuine cost from inflated claims

A scheme under pressure generates both, and they need treating differently. Genuine additional cost comes from instructed change, from design that was incomplete at tender, from conditions that differ from those assumed, and from delay the contractor is entitled to be paid for. That cost is real and arguing about it wastes the time that should be spent controlling the rest.

Inflated claims are a different matter: work claimed twice, change claimed that was always in the scope, preliminaries claimed for a period the contractor caused, dayworks where measured rates apply, or global claims for a sum with no linkage to the events that caused it.

Separating them is a matter of entitlement before value, and the test is documentary. What was instructed, by whom, with what authority, and what does the contract say about how it is valued. The method is set out in entitlement before value, and it is the same discipline whether the scheme is in difficulty or not.

Where records are thin on both sides, the position is weaker than either party believes, and the realistic outcome is a negotiated one. Recognising that early is cheaper than discovering it after a formal dispute has started.

Reset the control, not just the budget

A revised budget without revised control produces the same overrun again, a few months later. The reset is procedural: all instructions in writing and from one named person, every variation valued within the month it is instructed, no work proceeding on a verbal instruction, and a single cost report that both the contractor and the investor work from.

Authority is usually the root of it. On schemes that run away, several people have been giving directions on site, each of them reasonably, and the cumulative cost appears later as a list of claims. Closing that down is unpopular for a fortnight and saves the rest of the project, as we describe in who is allowed to instruct the contractor.

The contingency needs resetting too, honestly. If what remains is smaller than the risk still ahead, that is a funding conversation now rather than a surprise in three months. Where there is a facility, the lender will reach the same conclusion through its own monitoring, which is better met with your own figures than theirs.

Decide what the scheme is now

Once the position is clear, the commercial question can be answered: finish as designed, finish to a revised specification, or change the exit. Each has a number attached and the numbers are now reliable enough to choose between.

Value engineering at this stage is narrower than it was at design, because much of the scheme is built or committed. What is usually left is the fit-out specification, the external works and landscaping, and the phasing of the remaining packages. Cutting in the wrong place here damages the sale, so the test is whether the saving survives the valuation of the finished units.

Changing the exit is sometimes the right answer. A scheme that no longer works as a sale may work as a hold and refinance, or the other way round. That is a different appraisal rather than a cost exercise, and it needs the same rigour as the original one, which is the argument made in how much should you pay for a development site.

What good looks like after the reset

One monthly document containing the measured value of work in place, the status of every variation, the forecast final cost with the movement explained, the contingency position, the programme and the risks. Everyone working from it. Nothing significant discovered at the end that was not visible in it.

That is not a sophisticated system, it is a routine, and the reason a scheme loses control is almost never that the routine failed. It is that there was no routine. Putting one in is ordinary cost control and variations work, reported in the format described in investor reporting.

Where the relationship with the contractor has already broken down, the same information is what gets the scheme to a settlement rather than to an adjudication, and the whole recovery sits inside development management.