Price to win

Price to win versus price to deliver: is the proposed margin defensible?

If the price someone thinks you need to win sits below the cost of delivering the work, the margin is not defensible. Separate revenue, direct cost, overhead, contingency and margin from markup, then test base and downside cases. The arithmetic is evidence for a second gate, not a rule that always says bid or no-bid. Part of the economics of bidding.

Published: 16 September 2026 · By Andre · Reviewed by Bill · selrob

Takeaways

Price to deliver is a cost stack plus a named margin convention. Price to win is an input, not a forecast of the award.
Margin is profit divided by revenue. Markup is profit divided by cost. They are not interchangeable.
Overhead absorption is a recovery target. Most of it is still there if you lose.
A calculation does not decide the bid. It feeds re-qualification at pricing.

What do the words mean?

Revenue is the comparison total at the rates you offer, for the quantities the pack tells you to extend, excluding provisionals unless they are instructed. Direct cost is labour, materials and subcontract attributable to that work. Overhead is the share of branch cost you choose to recover on the job. Contingency is an allowance for the uncertainty you have named, not a hidden margin. Net margin is (revenue − direct − overhead − contingency) / revenue. Markup on direct is (revenue − direct) / direct.

Price to deliver is the revenue at which that net margin meets the firm’s named floor. Price to win is a figure someone believes may be required to be competitive. It is not a competitor’s bid. It is not a prediction that the buyer will award at that number. Write it as an assumption with an owner.

Which overhead treatment are you using?

Name it. The model below absorbs 12 percent of direct cost on every priced line. That 12 percent is a recovery target. If the contract is not won, most of the underlying overhead is still there. A cash view that ignores overhead will always make a thin bid look better. Put both numbers on the page and say which one the gate is using.

Contingency in this model is 10 percent of schedule-of-rates direct, zero on the lump (scope treated as closed), and provisionals are left out of the base case. Recurring FM work where the lump is not actually closed is the thin-margin problem; the sensitivity below puts 10 percent extra lump direct back in as a downside, not as a hope.

Worked example

Cost build: Western Ranges hard FM, year one

Hypothetical. The buyer, tender and figures are invented to show the fields in use; they are not a client result. Quantities and line IDs are the schedule on reading the pricing schedule. Term is three years; figures are annual; no CPI. AUD.

Delivery-floor cost build, completed hypothetical example
Line	Qty	Direct unit (AUD)	Direct (AUD)	Overhead 12%	Contingency	Delivery-floor rate / lump	Revenue in base
LS-01	1 year	125,000	125,000	15,000	0	156,000	156,000
SOR-01	400 h	80	32,000	3,840	3,200	110 / h	44,000
SOR-02	80 h	100	8,000	960	800	135 / h	10,800
SOR-03	250 sets	40	10,000	1,200	1,000	54 / set	13,500
PI-01 / PI-02	Excluded		0	0	0	39,500 / 310	0
Base total			175,000	21,000	5,000		224,300

Cost stack $201,000. Net margin $23,300 / $224,300 = 10.4 percent. Markup on direct $49,300 / $175,000 = 28.2 percent. Contribution $49,300 / $224,300 = 22.0 percent.

Blank template

Blank cost build

Delivery-floor cost build, blank form
Line	Qty	Direct unit	Direct	Overhead	Contingency	Offer rate / lump	Revenue in this case


Total

Name the overhead percentage, the contingency rule, and whether provisionals are in the case. Write the price-to-win figure as an input with an owner.

Cases

Base, price-to-win input, and two downsides

Hypothetical. The buyer, tender and figures are invented to show the fields in use; they are not a client result. The price-to-win input is BD’s nominated comparison total of $205,000, built by cutting LS-01 from $156,000 to $136,700 and holding unit rates. It is not a competitor price.

Margin cases, completed hypothetical example
Case	Revenue (AUD)	Direct	Overhead	Contingency	Stack	Profit	Net margin
Base, delivery-floor rates, indicative qtys	224,300	175,000	21,000	5,000	201,000	23,300	10.4%
Price-to-win input $205,000, indicative qtys	205,000	175,000	21,000	5,000	201,000	4,000	2.0%
PTW input, SOR volumes −30%	184,510	160,000	19,200	3,500	182,700	1,810	1.0%
PTW input, volumes −30%, lump direct +10%	184,510	172,500	20,700	3,500	196,700	−12,190	−6.6%

Volume −30 percent uses 280 / 56 / 175 against the B18 indicative 400 / 80 / 250. Lump +10 percent is $12,500 extra direct on LS-01 with no extra revenue. If overhead is treated as non-incremental, the PTW base case still shows $25,000 after direct and contingency ($205,000 − $175,000 − $5,000). That cash view is a treatment choice. It is not the 10.4 percent floor.

What happens if the winning price sits below delivery?

The margin is then a statement of hope. In the table, BD’s $205,000 input leaves 2.0 percent net at the volumes the pack printed, and a loss once indicative rates under-run and the lump is not closed. The model does not say no-bid by itself. It says the first call’s 10 percent net no longer holds. Take that sentence to re-qualifying at pricing. Finance owns the veto on this axis.

Do not invent a competitor’s number to justify the cut. Do not multiply $205,000 by three and call it a three-year bid. Do not treat provisionals as the recovery plan. The mix-shift case on the pricing-schedule page moves the comparison total by only $2,000 while doubling after-hours hours; the quality of the hours is the point, not the add-up.

What can this calculation not do?

It cannot predict the award. It cannot turn an uncertain price-to-win input into intelligence. It cannot price abatement, delay damages or an uncapped indemnity; those sit on the risk axis. It cannot make a thin bid good by renaming markup as margin. Put the cases on the record. Then make the call with named owners, not with the workbook.

Keep reading

Elsewhere in this guide.

Book a call

Bring the last tender you regretted.

In 30 minutes we’ll show you the call selrob would have made, and the working behind it.

Book a call