The real cost to tender a construction job

One failed bid, itemised

A subcontractor we spoke to priced a major fit-out tender. The pack was the usual weight: architectural drawings, a specification, preliminaries, and a pricing document that wanted quantities the drawings did not hand over. The firm did what most firms its size do when the measuring is beyond the week's spare hours — it sent the bill of quantities work out to an external quantity surveyor, and paid £1,200+VAT for the measure and the priced bill.

The bid was submitted on time. It lost. Nothing about that is a scandal — most bids lose, that is what a competitive tender is — but look at where the money went. The fee was not contingent on winning. There was no partial refund for a strong second place. The £1,200+VAT bought a document whose entire value expired the moment another firm's number was accepted, and the next tender that landed started the meter again from zero.

What a tender actually costs

Put the failed bid above into its three parts and you have the honest cost model for tendering. Every firm pays all three on every bid, whether it notices or not.

• Direct fees. If the take-off and the bill go to an external QS or estimator, that is an invoice per tender — and it arrives win or lose. Firms that keep the work in-house do not escape this line; they just pay it in salary or in the owner's unpaid time instead of in fees.
• Time. Measuring drawings, chasing the queries the pack leaves open, building rates, formatting the pricing document the way the main contractor wants it — for an owner-led firm this is evening and weekend work, done tired, after the paying jobs have had the daylight hours. Tired measuring is also where quantity errors are born, and a quantity error survives into the contract if you win.
• Opportunity cost. The least visible line and usually the largest. Every week the owner spends pricing one tender is a week of site supervision, client visits and quoting that did not happen — and every tender declined for lack of pricing capacity is a contract handed to a competitor without a fight.

The win-rate maths is the brutal part

Tendering economics punish the slow, expensive bid twice over. Suppose your firm wins one tender in four — a respectable rate in most markets. Then the true cost of the job you win carries the cost of the three bids that lost. If each bid costs you a fee of the size above plus a week of evenings, the winning job starts life carrying several thousand pounds of dead pricing cost before a single operative reaches site. Firms feel this without naming it: it shows up as margin that looks thinner than the tender said it would be.

The same maths runs the other way, and this is the part worth acting on. Win rate is partly out of your hands — you cannot control who else bids — but the number of tenders you price is not. A firm that can only price one tender a month is betting the quarter on three attempts. A firm that can price six is not. The constraint that keeps most subcontractors at one or two bids a month is not the market; it is pricing capacity.

Cutting the marginal cost of a bid

You cannot make tendering free, and you should not want the parts that need your judgement — margin, risk, whether the client is worth working for — to be automated away. What can fall to near zero is the marginal cost of the mechanical work: reading the pack, measuring the take-offs, structuring the bill to NRM2 and pricing it from your own rates. That is the work the £1,200+VAT paid for, and it is the work AI tendering and estimating now does in-house on a flat monthly plan, with the priced bill coming back to you for review the same day the pack arrives.

When the mechanical cost per bid collapses, three things change in practice. First, the tenders you used to decline for lack of time become worth an hour of review — your bid rate rises without anyone working later. Second, a lost bid stops being an invoice you cannot recover; it costs the review hour and nothing else, which takes the sting out of the win-rate maths above. Third, the owner's evenings come back, which sounds soft until you price what the owner's attention is worth on site and in front of clients. The quality of each bid rises too, because a measured, checked bill produced with a fresh head beats a midnight estimate every time — and because the documentation burden of JCT and NEC packages stops filtering you out of the bigger tenders before price is even discussed.

None of this requires hiring an estimator, and it does not turn pricing into a black box. The measure is itemised so you can check any line against the drawing it came from — the same discipline explained on the tendering and estimating page — and nothing is submitted until you have read it. If your pricing bottleneck is further upstream, in the measuring itself, start with take-offs and bills of quantities instead.

The question to ask about your last lost bid

Take the last tender your firm lost and write down what it cost: the fee if there was one, the hours at an honest rate if there was not, and the work that did not get priced while it was being put together. Then ask what your bidding would look like if that number were a tenth of the size. For most subcontractors the answer is not "we would spend less" — it is "we would bid on three times as much work", and that is the answer that grows a firm.