You just signed the biggest job the business has ever taken. It is a good number and you want it. But you have to pay people and buy material for weeks before the first check comes in, and you do not actually know if there is enough cash to get from here to there. "Can I afford to deliver this contract" is the question, and the answers you keep getting are a line of credit you have not read the terms of, a factoring quote, or a rule of thumb about how many months of expenses to hold. None of those is your number. A lender tells you what they will advance; a rule of thumb tells you what somebody else's business held. Neither tells you the day this job puts your account below the floor, or the deposit that would stop it.
This page is the large-contract decision from the owner's side of the desk, in the order the tests have to run. First, the cash curve for the job on the dates money actually moves, run twice: once on the terms as written and once on how this customer actually pays. Then that curve against the cash you hold and the fixed outflow the business keeps paying while it waits. Then the four other ways a good contract breaks a business: people you do not have yet, premiums that eat the margin, one customer becoming most of your revenue, and the work this job pushes out. It ends in one of four answers -- take it, take it with restructured terms and the specific ask named, take it only with financing sized and committed before you sign, or decline -- and the free large-contract cash calculator runs the same math on your numbers with every field blank until you fill it.
The Real Problem: The Risk Is Not Winning the Job, It Is Funding the Gap Between Doing It and Being Paid
Two mistakes sit under almost every large-contract cash scare, and both are made before the first invoice goes out.
The owner treats the risk as winning the work. Once the award letter arrives the hard part feels done, so the cash question gets a glance at the bank balance and a guess. But a contract is a promise to spend before you are paid. Material lands on day 10, payroll lands every two weeks whether or not the customer has paid, and the first progress billing does not become cash until the customer's own payment cycle has run. The exposure is the gap between doing the work and being paid for it, and on a job that is bigger than anything you have run before, that gap is bigger than anything you have carried before. The question is not whether the job is profitable. The worked example below is profitable by $80,000 and still cannot be delivered on the terms as signed.
The owner does the cash math on the contract's stated terms. Net 30 goes into the spreadsheet as 30. Then the first check arrives on day 45 because that is how this customer pays everybody, and the spreadsheet was wrong by two weeks on every receipt in the job. Stated terms are not behaviour. Your AR aging, or the last three invoices to this customer and the dates the money landed, is the input that makes the curve true; the contract number is what you use only when you have nothing else, and the calculator says so when that is what it is running on. If the terms themselves are the question, net 60 instead of net 30 or a discount for paying faster, what do these payment terms actually cost me prices them on this customer's real days-to-pay before the curve is run.
There is a third, quieter one. Financing gets assumed. "I have a line I have never fully used and a supplier who would probably give me terms" is two guesses dressed as a plan. A line counts when it is committed in writing with conditions you can name; supplier terms count when they are in writing from the supplier. Everything else is zero until it is not.
Can I Afford to Deliver This Contract? The Eleven Numbers That Decide It
These are the inputs. Every one comes off a document already in your hands: the contract, the bid worksheet, the bank statement, the payroll calendar, the line agreement. No accountant is required to produce them. The middle column says where each one comes from; the right-hand column says what happens if you leave it blank. A blank field is never filled in for you; where the page cannot run without a number you cannot produce, the table of answers further down names who can produce it.
| Input | Where the number comes from | If it is blank |
|---|---|---|
| 1. Contract value and the billing schedule: the dates and amounts you can invoice, or a single invoice on completion; the completion date; any deposit offered and whether it is kept on cancellation or recouped against later billings; whether the customer can terminate for convenience and on what notice | The contract or purchase order in hand | The one input the page cannot proceed without. No curve, no answer. |
| 2. Payment terms as written: net days, retainage percent and when it is released, whether retainage survives a termination | The contract | Terms blank is read as due on completion. Retainage blank is zero. No industry norm is substituted. |
| 3. How long this customer actually takes to pay | AR aging, or the last three invoices to this customer and the dates the money landed on the bank statement | The curve runs on the contract's net days and the page flags that it is running on the contract number, not behaviour. |
| 4. Direct cost to deliver, line by line, with the date each is paid: labor, material, subcontract, other; and the contribution this job must clear for you to want it | Bid worksheet or estimate; payroll calendar; supplier terms | Fewer than three dated cost lines and the cash tests cannot run; the page routes you to the job-costing and books diagnostics rather than guessing. |
| 5. Cash in the operating account today, the monthly fixed outflow the business pays regardless of this job, and the floor you will not let the balance go below | Today's balance; the last three bank statements | The buffer test and the concentration test cannot run. |
| 6. Committed credit: line limit, amount drawn, whether it is committed in writing, whether you know the conditions on it (covenant, borrowing base, clean-up, personal guarantee) and whether this job touches one | The line agreement or the last bank statement | Available credit is zero. A draw is never assumed. |
| 7. Whether the job needs people you do not have: when the shortfall starts, hours short, the day a hire would be productive, how many, loaded daily cost, payroll start and period | Payroll; your own hiring experience | No hire is modelled. A blank loaded cost blocks the affordability test and the page says so; it is never built from a published rate. |
| 8. The overtime, expediting or subcontract premium you would actually pay to hit the date without hiring: hours, rates, whether the premium form is decided, a subcontract bid you already hold | Your own rates; a bid already in hand, not one solicited for this | The margin-erosion test reports itself untested. Both overtime forms are shown when you have not said which. |
| 9. What share of revenue this customer will be after award (or trailing-twelve-month revenue and this contract's revenue), the notice period, and how long you would keep spending after a notice | Sales by customer; the contract | The concentration test cannot run. No share is estimated. |
| 10. The contribution the same crew and capacity would have earned from the work this job pushes out | Your backlog and past-job margins | Displaced contribution is not tested and the page says so. It is never read as zero. |
| 11. Your own cost of money: the line rate, or what the cash would otherwise earn | The line agreement, or your own judgment | Context only. Exposure is reported in days and dollars, never as an interest cost. |
| Plus one bound: the largest deposit or mobilisation payment this customer would realistically grant, as a percent of contract value | Your own knowledge of the customer | The deposit lever is not decided and no verdict rests on it. The field carries a provisional reference of 25 percent of contract value with its source stated as none; it is not a benchmark and it is not filled in. |
No industry, no revenue band, no business type is an input. Those change how the questions are phrased (retainage and progress billing for a contractor, deposits and retainers for an agency, work-in-progress lag for a professional firm) and nothing else.
Working Capital for Contractors and Everyone Else: The Cash Curve, Run Twice
The first test builds the job's own cash curve: every cost on the day it is paid, every receipt on the day it is expected to land, which is the billing date plus the days this customer takes to pay, with retainage held back until its release date. The curve starts at zero on the day you commit and ends at the job's contribution when the last dollar lands. In between it goes as low as it goes, and that low point -- the trough -- in dollars and on its date, with the number of days the curve sits below zero, is the first thing the page reports. If the curve never goes negative, the job funds itself on the schedule as written and the rest of the cash tests do not run.
It runs twice. The headline run uses your observed days-to-pay. The comparison run uses the contract's net days. The difference between the two is your own stated-terms-versus-behaviour number: how much deeper, how many days later, and how many more days exposed the real job is than the one on paper. No industry days-to-pay figure is placed beside it, because the number that matters is this customer's, and you already have it.
Two things feed back into the curve before it is tested. If the job needs a hire, the payroll from the day the hire starts is added to the outlays and the curve is re-run, and the page reports separately whether the hire lands after the schedule needs it and whether the business can carry the new payroll assuming this contract's receipts arrive on the customer's timeline, not on hire day. If hitting the date means overtime, expediting or a subcontract, those premiums are dated outlays too, and they thin the contribution at the same time they deepen the trough.
The Cash Flow Gap Against Your Buffer: The Position Curve
The job curve is then laid on top of the business. Position on any day is the cash you hold today, less the fixed outflow that has run by that day, plus the job curve, less the floor you set. The first day position goes negative is reported in its month; the lowest point, measured on the day before each receipt lands as well as on the event days themselves, is the cash shortfall: the headline number, on cash alone. Then, separately, the shortfall after the credit you treat as committed, so you can see how much of the plan is borrowed. A line that exceeds what is undrawn, a line that is not committed, a condition you cannot name or a condition this job would touch each raise a flag beside that number. No figure about how often businesses get the financing they apply for is placed beside the flags; the posture is simpler than that. Financing is secured before you sign or it is not in the plan.
Two tests run on the contract as a change to your book, whether or not the curve is funded. Concentration: this customer's share of revenue after award is reported as a fact, and the verdict comes from a different question -- if the customer gave notice on the day it could, and you kept spending through the period you have committed to, would the business go through its floor before the work could be replaced? A cancellable contract that only recovers on receipts the customer can stop is a compound exposure, and the test says so. No safe-concentration percentage is asserted anywhere on this page. Displaced contribution: the margin this job earns against the margin the same crew would have earned from the work it pushes out. Owners rarely compute it, and it is one of the four things that can decline a job on its own.
The Four Answers, in the Order They Are Tested
The order is load-bearing. The two answers that cost you nothing are tested before the one that costs money, and the one that costs money is tested before the job is declined.
| Answer | What has to be true |
|---|---|
| 1. Take it | The plan as signed never puts the business below its floor, the job clears the contribution you said it must, no decline trigger fires, and any hire the job needs lands in time. |
| 2. Take it with restructured terms | Some owner-side change to the deal closes the cash shortfall: a deposit or mobilisation payment, earlier milestones, shorter terms, less retainage or an earlier release, a price adjustment, or a start-date change where the billing dates are fixed to the calendar. The page names the specific ask and its size, in dollars, and states whether the concentration test still passes under it. It does not judge whether your customer will agree; that is your negotiation. |
| 3. Take it only with financing sized and committed in advance | No owner-side change closes it, but the shortfall is a number a committed line or a written arrangement can cover, and the credit flags are clear or can be cleared in writing before you sign. Sized two ways: the shortfall on cash alone, and the part beyond the line you already hold. A job with no margin never reaches this answer; you do not finance a job that cannot pay for the financing. |
| 4. Decline | Any one of four triggers fires and nothing cures it: the trough cannot be funded even after restructuring; the concentration is unsurvivable; the work this job displaces earned more than this job does; or the schedule cannot be met with the people the business can actually have. |
When a test the answer depends on could not run -- a blank that matters, a cost line without a date, a deposit bound you did not set -- the page emits no answer at all. It reports the cash result, names the one question that is holding the verdict, and stops. It never fills the blank with a guess to get to a verdict.
The six levers, and how each is sized. When the shortfall exists, each of six owner-side changes is tested on its own and the minimum that closes the hole is reported, or the residual it leaves. A deposit shifts the whole curve up by its amount, so the deposit that closes the shortfall is the shortfall itself when it is not recouped against later billings, and it is found by search when it is. Moving milestones earlier and shortening terms both move receipts earlier, which cannot help on the days before the first billing, and the page says so when that is where the low point sits. Retainage is tested as a lower percent and an earlier release. A price increase only helps after money starts landing, so it cannot close a hole that sits before the first receipt. A later start only helps when the billing dates are fixed to the calendar; if billing follows the work, the whole curve moves together and nothing is gained, and the page asks you which it is rather than assuming. The levers are shown grouped by what the arithmetic found -- closes alone, closes only in combination, cannot close alone -- and never ranked by which one you should ask for. A combination is reported when you select it, and only then.
The deposit lever is bounded by you. An unbounded deposit closes every shortfall on paper, which makes it an escape hatch rather than a lever. So the page asks for the largest deposit this customer would realistically grant, as a percent of contract value, and tests the deposit against it. If the deposit that closes the hole is inside the bound, the lever closes it. If it is above the bound, the lever cannot close it alone, the residual is stated, and the page reports which combinations of a deposit at the bound plus another lever would, for you to select. If you leave the bound blank, the deposit lever is not decided, and if nothing else closes the shortfall the page stops with that one question rather than assuming a deposit you have not said you can get.
Worked example: the $200,000 contract on net 30 that pays in 45
A $200,000 contract, billed in three progress invoices: $70,000 on day 30, $70,000 on day 60, $60,000 on day 90. Net 30 as written, no retainage, no deposit offered. The owner's last three invoices to this customer landed in 45 days, so the headline run uses 45. Direct cost is $120,000: $30,000 of material on day 10, $20,000 of labor on each of days 14, 28, 42 and 56, and $10,000 of other cost on day 70. The owner wants the job to clear $50,000. Cash in the account is $40,000, fixed outflow is $6,000 a month, and the floor is $5,000. A $60,000 line is undrawn, committed in writing, with conditions the owner can name and none this job touches. No hire, no premium. This customer would be 30 percent of revenue after award; the contract can be cancelled on 45 days notice, and the owner would keep spending for 20 days after a notice. The work it displaces would have earned $30,000. The contract says a mobilisation payment, if one were granted, would be kept on cancellation and would not be recouped against later billings.
| Test | Result |
|---|---|
| Job cash curve, on how the customer pays (45 days) | Receipts land on days 75, 105 and 135. Trough $120,000 on day 70. Negative from day 10 through day 104: 95 days exposed. |
| Job cash curve, on the terms as written (30 days) | Trough $110,000 on day 56; 80 days exposed. |
| The two-run gap | $10,000 deeper, 14 days later, 15 more days exposed. This is the owner's own stated-terms-versus-behaviour number. |
| Position against the buffer | First below the floor on day 14, in month 1. Lowest point $99,794.52 below the floor, the instant before the first receipt lands on day 75. Cash shortfall: $99,794.52. |
| After committed credit | $60,000 of the hole can be carried on the line: shortfall after credit $39,794.52. Half of the job's own trough would be borrowed. The line is smaller than the shortfall, so the credit test flags it. |
| Contribution | $200,000 less $120,000 = $80,000, above the $50,000 the owner requires. The job is profitable. It still cannot be delivered as signed. |
| Displaced work | $30,000 displaced against $80,000 earned. Does not fire. |
| Concentration, on the plan as signed | If notice came on day 45, every receipt is after the notice and is removed, and spending runs to day 65. The floor is breached on day 14 and the low point in that window is $87,821.92 below it. Unsurvivable as signed, for the same reason the cash test fails. |
| Levers | A mobilisation payment of $99,794.52 at signing (49.90 percent of contract value), not recouped, closes the shortfall exactly. A price increase cannot close it: the low point sits before any money lands. Retainage is not available: there is none to reduce. Earlier milestones and shorter terms each cannot close it alone: even a first invoice on day 0 paid in 45 days lands after the day-42 payroll. |
| Concentration, under the deposit | With the $99,794.52 kept on cancellation, the same notice on day 45 leaves the business $11,972.60 above its floor at the worst point in the window. No breach. |
| Answer | Take it with restructured terms: a mobilisation payment of $99,794.52 at signing, kept on cancellation. Whether the customer grants it is the owner's negotiation. The concentration test covered the notice period plus committed spend; it did not cover how long the business would take to replace the work. |
Now the bound. If this owner enters 50 percent or more as the largest deposit the customer would grant, the deposit lever closes the shortfall and the answer above stands. If the owner enters 25 percent, the deposit is capped at $50,000, the lever cannot close the hole alone and leaves $49,794.52 uncovered, and the page reports that a deposit at the bound combined with shorter terms would close it, for the owner to select; it selects nothing. If the owner leaves the bound blank, the page emits no answer, reports the $99,794.52 shortfall and the deposit that would close it, and asks the one question: whether that deposit is a real lever or an escape hatch depends on the bound left blank. Same job, same numbers, three different states, and the difference between them is a fact about the customer that only the owner has.
What to Do With Each Answer
| Answer | What it means | Next action |
|---|---|---|
| Take it | The job funds itself on the schedule as written and nothing else fires. | Sign. Put the receipt dates from the behaviour run, not the terms run, into a 13-week cash flow forecast so the first late check is visible the week it is late, not the month after. |
| Take it with restructured terms | A specific ask closes the hole. The page has sized it. | Ask for the named change before you sign, in the contract, not by email. If the customer will not grant the full ask, re-run with what they will grant as the bound and see whether a combination closes it. A change that costs you nothing is tested before one that costs interest. |
| Take it only with financing sized and committed in advance | No owner-side change closes it; the number is financeable. | Get the commitment in writing, with the conditions, for at least the shortfall after credit, before signing. Read the covenant. A line you have not read counts as zero. If the lender wants financials you do not trust, that is a separate problem and it comes first. |
| Decline | One of the four triggers fires and no lever cures it. | Reprice, restructure the schedule, or walk away. If the trigger was displaced work, the number to reprice to is the contribution the displaced work would have earned. If the trigger was concentration, a cancellation fee or a deposit kept on cancellation is the ask that changes the test. |
| No answer, one question | A test the answer depends on could not run. | Answer the question the page named. If it is a cost line you cannot date or a days-to-pay you cannot produce, the making-money-on-each-job diagnostic and who should help with my books name what would let you produce it. A controller or a fractional finance person is one way to get those numbers produced; it is never the answer this page gives. |
Where the Numbers Come From, and Which Ones Are Yours
Every number that decides the answer is yours: the contract, the bid, the bank statement, the payroll calendar, the line agreement, and your own knowledge of how this customer pays and what they would grant. That is the point, and it is also the boundary.
- No benchmark of any kind. No typical retainage, no industry days-to-pay, no months-of-expenses rule, no buffer figure from a survey, no safe-concentration percentage. Where you want to know whether your number is normal, the page answers with your number and the arithmetic on it, and nothing else.
- No probability of financing, and no financing product. The page sizes the commitment you would need and treats it as absent until it is in writing. It does not estimate approval odds, it does not offer factoring, and no path on it ends at a lender, a factor or a provider.
- No reading of your contract. Termination for convenience, retainage release, whether a deposit is kept on cancellation, covenants, bonding: the page asks what the document says and scores the plan on your answer. What the clause means is a question for the contract's own text and your attorney.
- The federal figures beside the loaded-cost field are context, not inputs. Employer Social Security and Medicare rates and the 2026 wage base, the net federal unemployment rate on the first $7,000 per employee, and the overtime rule for non-exempt workers are shown beside the field where you enter what a hire would cost you per day. They are never placed inside a field and never used to build a number you did not enter; a blank loaded cost blocks the hiring affordability test and says so. State unemployment tax is per state and is not included.
- No first-party figure without its denominator. Nothing on this page reports what other owners entered, how many took the job, or how often a deposit was granted. What you enter is not aggregated into a benchmark for anyone else.
- Every dollar figure on this page is the worked example's own arithmetic, reproduced by the same script the calculator serves. The 25 percent shown beside the deposit-bound field is a provisional reference with its source stated as none; it is not filled in and it is not treated as a fact.
When This Is Not the Right Starting Point
If you cannot date your direct costs or produce how long this customer takes to pay, the curve is a guess and so is everything after it; job costing help for contractors is the page for the cost side, and not trusting your numbers for the rest. If the contract is already delivered and the question is where the cash went, that is the same symptom after the fact rather than before it: profitable but no cash routes the retrospective version, and why profitable, no cash names the seven usual drains. If the thing that made you nervous is that this one customer will be most of what you do, the concentration test on this page is the survivability question, and customer concentration and revenue quality covers what a concentrated book does to the business over time, not just through this job's notice period. If the job needs people you do not have, the hire is its own decision with its own loaded cost and its own runway test, and can you afford another employee runs it; this page then adds that payroll to the curve. If you are a contractor and jobs look fine at bid and finish thin, the leak is usually tracking, and why your jobs look profitable until the final invoice is the page for it. And if revenue is climbing while the balance shrinks across every job, not this one, growing too fast and out of cash is the version of this problem that is already happening.
Frequently Asked Questions
Can I afford to deliver this contract?
You can if the cash curve for the job, built from your own billing dates, your own cost dates and the days this customer actually takes to pay, never puts the business below the cash floor you set, and no decline trigger fires. If the curve does go below the floor, the answer is one of three: an owner-side change to the terms closes the hole (a deposit, earlier milestones, shorter terms, retainage, price or start date), financing sized to the hole and committed in writing before you sign closes it, or nothing closes it and the job is declined. The calculator runs those tests in that order on your numbers and names the specific ask when one closes it.
How much working capital does a contractor need for a large job?
The amount the job's cash curve goes below zero at its lowest point, measured on the dates receipts actually land rather than the dates the contract says they are due, plus the fixed outflow the business keeps paying while it waits, less the cash you hold above your own floor. There is no percentage of contract value that answers this, because two jobs of the same size with different billing dates and different payment behaviour need different amounts. The worked example on this page is a $200,000 contract that needs $99,794.52 more than the business holds, and a $200,000 contract billed differently would need something else.
Should I ask for a deposit or mobilisation payment before signing?
If the cash shortfall exists, a deposit at signing is the first lever tested, because it costs the business nothing and it is the only lever that helps on every day of the job. The calculator states the deposit that closes the shortfall to the dollar. Whether that deposit is one your customer would grant is your call: you enter the largest deposit you believe this customer would realistically agree to, as a percent of contract value, and the deposit lever is tested against that bound. Left blank, the deposit lever is not decided and no verdict rests on it.
What if the contract pays net 30 but the customer actually pays in 45 days?
Then the cash curve runs on 45, not 30, and the difference between the two runs is reported as its own number. In the worked example the two-run gap is $10,000 deeper, 14 days later and 15 more days exposed. Stated terms are not behaviour; the last three invoices to this customer and the dates the money landed are the input, and if you do not have them the calculator says the curve is running on the contract number and flags it.
Does this tell me whether a bank will lend me the money?
No. Financing is the third answer, tested only after the owner-side changes fail, and the calculator sizes it: the shortfall on cash alone, and separately the part that is not covered by a line you already hold. It does not estimate whether a lender would approve it, and it treats a line as available only when it is committed in writing with conditions you can name. A line you have never drawn with a covenant you have not read counts as zero until you have read it.
Sources
- Internal Revenue Service, Tax Topic 751, Social Security and Medicare Withholding Rates: employer share 6.2 percent and 1.45 percent; 2026 Social Security wage base $184,500; Additional Medicare Tax 0.9 percent is employee-only. Page last reviewed or updated 2026-01-20; read 2026-09-13. Shown as context beside the loaded-cost field only.
- Internal Revenue Service, Tax Topic 759, Form 940, Employer's Annual Federal Unemployment (FUTA) Tax Return: 6.0 percent on the first $7,000 of wages per employee; maximum credit 5.4 percent; net rate after credit 0.6 percent, higher in credit-reduction states. Page last reviewed or updated 2026-02-25; read 2026-09-13. Shown as context beside the loaded-cost field only.
- 29 U.S.C. 207(a)(1), Fair Labor Standards Act: overtime at not less than one and one-half times the regular rate for hours above forty in a workweek for non-exempt employees. Read 2026-09-13 at the Office of the Law Revision Counsel. Shown as context beside the overtime inputs; the premium the page uses is the one you enter.
- Every other number on this page is the worked example's own arithmetic or comes from your records. No days-to-pay, no retainage, no cash reserve, no concentration percentage and no financing figure on this page is a benchmark.
Next Step
If the question in your head is whether you can deliver the job you just won, the first move is the cash curve on the dates money actually moves, before the line is drawn and before the first payroll lands. The large-contract cash calculator at GetAFractional runs it in about ten minutes: the trough and the days exposed on how the customer pays and on the terms as written; the cash shortfall against your own floor and what is left after the credit you hold; the hiring, premium, concentration and displaced-work tests; the six levers with the minimum ask that closes the hole, bounded by the deposit you say the customer would grant; and the answer -- take, take with a named ask, take only with financing committed first, or decline -- with every field blank until you fill it.
If the page stops on a question you cannot answer from your books, the making-money-on-each-job diagnostic sorts out whether the cost side is a tracking problem, and who should help with my books names the role that can produce the numbers. If the contract is already delivered and the cash is already gone, the profitable-but-out-of-cash diagnostic is the retrospective version of this page.
This article is informational and does not constitute financial, legal, or tax advice. Consult a qualified professional for decisions specific to your situation.