The Construction Cash Flow Crisis in Numbers: What 21.4 Days of Cash and 82% Payment Delays Measure

21.4 Days

RSM’s discussion of the CFMA 2020 survey reports 21.4 days of cash on hand. It is a historical average, not a current Level measurement or a median for service contractors. If your P&L looks healthy and your bank balance does not, the four measurable gaps between the two are broken down in Profitable on Paper, Broke in the Bank. Not 21 weeks. Not 21 months. Days of cash is cash divided by average daily operating cost, so 21 days is a rough cushion against outflows, not an exact countdown to a missed payroll; receipts keep arriving and payables timing matters too.

Rabbet’s 2024 report estimates a $280B cost of slow payments and reports 82% of respondents experiencing payment delays longer than 30 days. That is a modeled cost and a respondent share, respectively. Neither is the value of unpaid receivables, the share of all invoices paid late or evidence that slow payment caused business closures.

The primary Rabbet PDF was recovered and checked October 7, 2026. Its study page describes an August 2024 online survey; its trends page gives 82% for 2024 versus 49% for 2022. Those are respondent shares from different survey years, not the fraction of dollars or invoices paid late.

This article separates three evidence types: historical association and vendor surveys, a reproducible public-company receivables calculation, and Level’s contractor metrics. Their populations and periods differ. The operator question is which mechanism your own ledger reveals: delayed billing, a rejected pay application, an agreed holdback or a late customer payment.

DSO: How Long You Actually Wait

Days Sales Outstanding (DSO) is usually calculated as accounts receivable divided by revenue for a period, multiplied by the days in that period. It is a balance-based average, not the wait on any single invoice, and it moves with retainage, billing timing and whether contract assets are included. The rows below use different definitions, so compare them by mechanism, not as one ranked list:

The public-company calculation above uses year-end receivables divided by full-year revenue. The CFMA respondent measure and Billd’s self-reported payment cycle are not the same calculation. The CFMA figures are supported by its official indexed highlights; direct PDF access was blocked during this review. Separate trade receivables, reported retainage and contract assets, and compare comparable billing models. A 60-day review trigger is an owner planning choice, not a measured cutoff for every trade. Contractual notice, prompt-payment, retainage and lien deadlines control the next action.

For a deeper breakdown of what DSO means for your specific situation, see DSO: The Number Every Contractor Should Know.

The Late Payment Epidemic

A balance-based DSO average does not describe every invoice or contractor's payment-delay experience. The survey share below measures respondents, not an invoice on-time payment rate.

• 82% of respondents experienced payment delays longer than 30 days in Rabbet’s August 2024 survey, versus 49% in its 2022 survey. Its respondent mix was 93% general contractors and 7% subcontractors; it does not publish a respondent count we could verify

Do not combine percentages from different survey populations into a single contractor late-payment rate. In your own ledger, distinguish invoices not yet due, contract retainage, disputed change orders, approved but rejected pay applications and overdue accepted invoices. Those buckets require different actions.

Billd’s 2026 report describes a 51-day average payment cycle and says 64% of subcontractors reported being slow paid. Its earlier 2025 web edition currently gives 54 days, while the previously cited report PDF gave 56 days and is now unavailable at its old URL. Use the named edition; the difference does not establish that your customer pays faster or slower. Ask who starts the clock, when the pay application was accepted and when funds actually cleared.

$280 Billion, And Growing

Rabbet, a construction finance software vendor, has published annual estimates of the cost of slow payment in construction since 2020. These are modeled vendor estimates, not audited totals, and the series below has no 2023 entry:

Rabbet frames that $280 billion as roughly 14% added to total U.S. construction spending. Its scale arithmetic is $2.0 trillion of anticipated 2024 construction spending x 14% = $280 billion. This reproduces the publisher’s rounded projection, not an independently estimated causal cost. At the company level, look for identifiable costs: financing receivables, staff time chasing payment and bid cushions all show up in a contractor's own books. Smaller subs and specialty contractors with thin cash buffers have the least room to absorb it, though the estimate itself does not measure how the cost is distributed.

What 21.4 Days of Cash Actually Means

The 21.4-day figure comes from RSM’s discussion of the CFMA 2020 survey. Do not treat it as the present cash reserve of a $3-10M trades company. Calculate your own available unrestricted cash and map dated inflows and obligations. Survey context is useful only when its population and year are clear:

• Billd’s 2026 report reports 64% of subcontractors being slow paid. It is a vendor survey, not a census or a direct measurement of your required cash reserve.

Subtracting days of cash from DSO does not calculate a financing requirement. The first ratio uses cash and operating costs; the second uses receivables and sales. Payroll cadence, supplier terms, deposits, opening balances and the timing of actual receipts all matter. Build a weekly schedule of those dollar flows instead of subtracting unrelated averages. A fictional counterexample shows why. Both companies below have $21,400 cash, $1,000 average daily operating cost, $56,600 ending AR and $365,000 annual credit sales. Both show 21.4 cash days and 56.6 ending-AR DSO ($56,600 / $365,000 x 365). Assume $7,000 of cash outflows each week, no other receipts or payments, and no minimum cash reserve:

The arithmetic is $21,400 + receipts through week 5 - five weeks x $7,000. The ratios are identical; the dated receipt changes the financing need. Recalculate every week and add any separately justified liquidity reserve. Neither the 35.2-day difference between the ratios nor a survey percentage sizes the line of credit.

One documented vendor observation:

• In Rabbet’s 2024 survey, 97% of responding general contractors said they had charged developers more for projects in the prior year due to slow payments. This is a respondent statement, not a measured price increase across every project

For a bid, calculate the carrying cost from the cash actually advanced, the assumed financing rate and the period until payment. Keep that cost separate from any late-payment penalty, bad-debt allowance and target profit. The resulting dollar estimate is more useful than applying an unrelated survey’s percentage to your price.

What Our Data Shows

Those are the industry numbers. The figures below come from the Level Index, built on the Level founding team's historical analysis of 2,242 contractors representing $13.25B in job revenue and 2.5M invoices across operating, private-equity and CFO roles. Those are the broader research-universe figures; each metric below has its own sample and definition.

The collection rate gap is real, but it is not all recoverable cash. Collection rate here is cash collected divided by billed revenue, measured across 464 contractors. The median is 85.1%; the 90th percentile is 96%. Fictional arithmetic: at $10M in billed revenue, the 10.9-point gap between 85.1% and 96% is about $1.09 million of billings not converted to cash in the comparison window. That is not automatically collectible A/R, permanent loss or recurring profit. It can include invoices still within terms, retainage not yet due, credits, disputed amounts, write-offs and cutoff timing. Bridge it with invoice aging, credit memos and retainage schedules before deciding what is recoverable.

A large billed-versus-collected gap can come from retainage, credits, disputed changes, legitimate timing differences or collection failure. Reconcile the components before promising recovery. The underlying workflow is explained in Why Contractors Bill Millions and Collect Thousands Less.

Slow invoicing compounds the problem. The Level Index reports a median of 1 day from completion to invoice including progress billers, with a reported sample of 733; the adjusted post-completion median is 7 days, without a separately published subgroup count. Invoice-to-payment time is a separate interval. Fictional timing arithmetic: 7 days before invoicing plus 57 days after invoicing equals 64 days from finished work to cash. A balance-sheet DSO ratio is not a clock on a particular invoice. See How Fast Should You Invoice? for the full benchmarks.

And then there is logged work with no invoice linked in the extract. The historical review flagged about 288,000 job records across 1,779 contractor records. That is a review queue, not proof that no invoice was ever sent: warranty, callbacks, agreements, internal work and broken links can explain a missing match. Check eligibility, invoice identifiers and later billing before attaching a dollar value. The full breakdown is in 288,000 Jobs Where Technicians Logged Hours and Nobody Sent an Invoice.

Layer in retainage and contract-specific approval requirements. Track release eligibility, documents still required, the payer and any notice deadline separately. State law and the signed contract take precedence over an internal aging routine.

The Fix Is Process, Not Prayers

The collection-rate distribution does not identify why the top decile collects more; customer mix, contract terms and retainage all play a part. Process is the part an operator controls, and it can be measured.

Test AR automation on the mechanism it can change. A reminder cannot fix a rejected pay application, an incorrect customer entity or a missing closeout document. Tesorio’s current page advertises a 33-day average DSO reduction for its customers, rather than the 32% previously cited here. It is vendor marketing, with no matched contractor population or independently controlled comparison on that page. In a pilot, measure billing lag, acceptance date, reminder delivery, disputed dollars and cash received, using the same scope before and after.

Price the manual work from your own time log. Fictional example: 200 invoices a month x 15 minutes each x $40 an hour of loaded administrative cost = $2,000 a month, or $24,000 a year. This measures assumed time cost, not cash savings from automation. Keep review and exception-handling time in the comparison; reclaiming hours does not itself reduce payroll.

Practices worth testing against your own numbers: invoice within 24 hours of job completion (or progress-bill during the job), hold a weekly AR review, track DSO and collection rate by customer, and treat collections as a core business function, not an afterthought when cash gets tight. Measure billing lag and collection rate before and after each change so you know which one moved.

The Seasonal Multiplier

Everything above gets worse at predictable times. Across 217 contractors with more than $500K in revenue, the median company earns 47% of annual revenue in its 3 strongest months. If your own revenue is concentrated in Q4 and the relevant invoices are collected in Q1, payroll and restocking can come due before those receipts. The three strongest months in the benchmark need not be consecutive or fall in Q4. A thin cash buffer can run down, and a forecast can reveal a need for funding before payroll comes due.

---

The Bottom Line

The historical surveys document payment strain in their respondent groups. Level’s separately defined collection-rate metric reports a median of 85.1% across 464 contractor records and a 90th percentile of 96%. At an illustrative $10M billed, the 10.9-point comparison is $1.09M not converted to cash in the comparison window. It is not automatically an AR balance, permanent loss, recoverable cash or recurring profit.

The distribution is a reference point. The top decile in the collection-rate measurement collects 96% or more of billings. The data does not show that those companies share the same customers or terms as everyone else, so treat their result as a reference point, then work on what you control: billing lag, weekly AR follow-up, contract payment terms and visibility by customer.

Q: What's the single most impactful thing I can do for cash flow this month? A: Measure your actual DSO and collection rate. Choose ending or average eligible AR, divide by comparable period sales and multiply by the period’s days; document whether retainage, contract assets and credit sales are included. For collection rate, match receipts to the intended billing scope and reconcile cutoff, credits and write-offs. The published Level ratio does not disclose a measurement window, so it is a reference point rather than a matched comparison until you define your own. Fictional scale check: $5M annual credit sales / 365 x 10 fewer receivable days = about $136,986 of one-time balance reduction, if sales and scope stay constant. We calculate both in our free profitability audit.

Q: How does Level help with contractor cash flow? A: We use available connectors or reviewed exports from your QuickBooks and field service software, build an AR dashboard with documented source refresh timing, and run weekly reviews on your collection rate, DSO, billing speed, and aging buckets. We flag the specific customers, invoice categories, and process gaps that are costing you cash, then work with your team on fixes. We track DSO, collection rate and billing lag from your own baseline so you can see what changed. The first audit is free.

Q: Should I invest in AR automation software? A: Evaluate the workflow and the economics. Pick a representative set of accepted, disputed and not-yet-due invoices, record the baseline, then test whether the tool changes the actual bottleneck. Compare time saved after human review, collection timing and customer experience. Do not assume a vendor’s DSO percentage applies to your ledger. But software alone doesn't fix the problem. You need someone reviewing the data and acting on it weekly. That's the difference between a bookkeeper and a CFO.

Q: How do I protect cash flow against seasonal swings? A: Start with your seasonal revenue pattern. The benchmark’s roughly 47% describes the three strongest revenue months, which need not be consecutive or form a quarter. If your busiest months precede payroll or supplier payments and customers take 60-90 days to pay, receipts can arrive too late for those obligations. Use a 13-week forecast built from dated invoices, expected payments, payroll and supplier obligations. Level builds these where the signed engagement includes cash forecasting.

For the five-company receivables calculation, accounting scope and source reconciliation, see the full public-company DSO benchmarks. The historical survey figures are separate from Level’s contractor research. Define the scope of your own ledger before using a survey figure as a target.