Quoting a Job: Where the Margin Goes | HL Hunt
Quoting a Job: Where the Margin Goes
A quote is rarely wrong when it is written. It is wrong by the time the job finishes, and it gets there in increments too small to argue about. A job quoted at $8,400 with a planned 20% margin finished at 0.8%, because labour ran 18% over, materials came in 13% high, and three small changes went unbilled. The figure that should change how you write the next one: replacing that $1,612 takes another $8,060 of revenue, which is an entire second job to recover what the first gave away.
What you'll learn
Where a profitable quote goes
The quote is not usually wrong when it is written. It is wrong by the time the job finishes, and it gets there in small increments that each look too minor to bill for.
Take a job quoted at $8,400 against a planned cost of $6,720, leaving a planned margin of $1,680, or 20%. Then the ordinary things happen:
| What happened | Amount |
|---|---|
| Labour overran: 104 hours against 88 quoted, at $52 loaded | $832 |
| Materials came in over: $2,140 against $1,900 | $240 |
| Three small changes, none billed: roughly $180 each | $540 |
| Total erosion | $1,612 |
Now the figure that should govern how the next quote is written: to replace $1,612 of lost margin at a 20% margin, you need $8,060 of additional revenue. That is another whole job of the same size, won and delivered, to recover what one job gave away.
Per our pricing analysis, this is what makes margin erosion different from a pricing mistake. A job priced 5% too low loses 5%. A job priced correctly and delivered without controls lost 96% of its margin, and the loss is invisible until someone reconciles the job rather than the month.
The labour estimate is the whole problem
Materials are easy to estimate because they are countable. Hours are not, and hours are the larger number. Per our terms analysis, the same asymmetry applies on the payment side: what is countable gets controlled and what is judged gets away.
Three things cause hour overruns, and only one of them is estimating badly:
- Unquoted work performed anyway. The largest cause, and it is not an estimating error. It is a scope control failure.
- Optimism about the base case. Estimates are usually built on the job going normally, and jobs going normally is not the average case.
- Unbilled travel, setup, waiting, and rework. Time that is real, chargeable in principle, and left off the sheet because it does not feel like the work.
The fix for the first is contractual and the fix for the second is arithmetic. If your jobs historically run 18% over on hours, quoting at your estimate means quoting at a figure you exceed most of the time. Either quote the historical actual or quote the estimate plus the historical variance, and say which you are doing.
The number to compute before your next quote
Take your last ten completed jobs. For each, write down quoted hours and actual hours, and compute the ratio. The average of those ratios is your estimating factor, and it is almost never 1.0. If it is 1.18, then every estimate you produce should be multiplied by 1.18 before it becomes a price. This takes an afternoon and it is the single highest-return piece of analysis available to a job-based business.
Change orders, and why they go unbilled
Nobody fails to bill a $3,000 change. Everybody fails to bill a $180 one, and three of those was a third of the erosion above.
They go unbilled for predictable reasons:
- The amount feels too small to raise. Set against the relationship, $180 looks like a bad trade, which it is not when it happens three times per job across forty jobs a year.
- The change was requested informally. A conversation on site, with no paper, becomes a thing you did rather than a thing you sold.
- Nobody decided who can approve one. If the person on site cannot authorize a charge, the charge does not get made.
- The mechanism is slow. A change order process requiring an office, a form, and a signature will be bypassed by anyone trying to finish the job.
The fourth reason is the one that undoes the other three. A change order that takes two minutes on a phone, with a photographed signature or a texted confirmation, gets used. One that takes twenty minutes gets skipped, and skipping it is rational for the person on site.
The contract should state the rate for additional work before any additional work is requested, because agreeing the rate at the moment of the request is a negotiation from a weak position.
Writing a quote that holds
Six things, in order of how much they protect the margin.
- State what is excluded, not only what is included. Inclusion lists are read as illustrative. Exclusion lists are read as boundaries.
- Put the additional-work rate in the quote. An hourly or unit rate for anything outside scope, agreed in advance.
- Set an expiry date. A quote accepted five months later is priced on five-month-old material costs, and per our pricing analysis the cost of not doing this rises with input volatility.
- Name the assumptions the price depends on. Access, conditions, information supplied by the customer, decisions made by a date. Each is a thing that can change the cost and each should be identified as such.
- Specify payment timing against milestones, per our receivables analysis, because a quote is also a cash flow document and a job that is profitable and paid in 90 days can still break the business.
- Quantify the deposit against your exposure, not as a convention. If materials are $1,900 and non-returnable, a deposit below that number means you are financing someone else's project.
Item four is the least used and the most useful in a dispute. "Price assumes clear site access and that final specifications are provided by the 12th" converts a later argument about fault into an already-agreed condition.
The deposit is a credit decision
Taking a small deposit on a job with large up-front costs is extending unsecured credit to the customer, whether or not anyone describes it that way. Per our deposit analysis, the deposit should cover non-recoverable costs at minimum. Where it cannot, the correct response is to assess the customer's credit rather than to hope, because that is precisely what you are doing.
Measuring by job, not by month
Per our year-end analysis, a business that reconciles monthly knows whether it made money and not where. A business that reconciles by job knows which work to stop selling.
| Metric, per job | What it reveals |
|---|---|
| Quoted hours against actual | Your estimating factor, which belongs in every future quote |
| Change orders raised against changes made | The unbilled work rate, usually the largest single leak |
| Actual margin against quoted margin | Which job types are profitable, which is often not the ones people assume |
| Days to payment | Whether a profitable job is a financeable one, per our timing analysis |
| Quote-to-win rate by price band | Whether you are winning work because you are cheapest, which is a different business than you think you are running |
The last row catches a specific and expensive pattern. A business winning almost every quote is priced too low, and a business winning very few is either priced too high or quoting the wrong work. A win rate near 100% is not a sales success, it is a pricing signal, and per our rationing analysis the choice between raising the price and turning work away is the same choice in either direction.
The financing side of quoting well
Jobs with material costs up front and payment at completion consume cash even when every quote is perfect. HL Hunt Business Credit Builder establishes tradelines in your business's name, so when a facility is needed to bridge a job, the file a lender reads is your business's own record.
When to requote rather than absorb
Not every overrun should be billed, and pretending otherwise damages relationships that are worth more than the amount. The distinction that works:
- Absorb your own errors. Something you estimated wrong, or did twice because it was done wrong once, is yours.
- Bill changes in what was asked for. A different specification, an addition, a decision reversed. These are new work and should be priced as new work.
- Bill changes in the assumptions you named. Which is the reason to name them in the quote.
- Raise it immediately, not at the end. A change discussed on the day is a conversation. The same change presented in the final invoice is a dispute, and per our collection analysis a disputed invoice is a slow invoice regardless of who is right.
The timing point in the fourth bullet does most of the work. The same $540 raised on the days it arose is usually paid without comment. Presented as a surprise at the end, it invites a review of the whole invoice, delays the entire payment, and costs more in collection time than the amount itself.
Frequently asked questions
Because erosion arrives in small increments. An 18% labour overrun, 13% on materials, and three unbilled changes took a $1,680 margin to $68.
$1,612 of lost margin at a 20% margin needs $8,060 of new revenue to replace, which is another whole job of the same size.
Compute your estimating factor from the last ten jobs: average actual hours divided by quoted hours. If it is 1.18, multiply every estimate by 1.18.
Absorb your own errors; bill changes in scope and in your stated assumptions. Raise it the day it happens, because the same amount raised late becomes a dispute.
Key takeaways
- A $1,680 planned margin finished at $68 from an 18% labour overrun, 13% on materials, and three unbilled changes.
- Replacing $1,612 of lost margin at a 20% margin requires $8,060 of additional revenue.
- Compute your estimating factor from the last ten jobs and multiply every estimate by it before it becomes a price.
- Unquoted work performed anyway is a scope control failure, not an estimating error, and it is the largest cause of overruns.
- Small changes go unbilled because the process is slow; a two-minute change order gets used and a twenty-minute one gets skipped.
- State exclusions and name the assumptions the price depends on, because both convert later arguments into agreed conditions.
- Size the deposit against non-recoverable costs, since a smaller one is unsecured credit extended to the customer.
- Raise every change on the day it arises, because the same amount presented at the end delays the whole invoice.
This guide is educational and does not constitute financial, accounting, or legal advice. All figures are stylized illustrations using assumed labour rates, hours, material costs, and margins, and do not describe any particular business or trade. Contract terms, change order requirements, deposit rules, lien rights, and consumer contract protections vary by industry and jurisdiction. Have your customer agreement and quote template reviewed by a qualified attorney before relying on them.