Profitable on Paper, Broke in the Bank: The Four Gaps That Eat a Contractor's Cash

Run the Cash Bridge

Use the same period and accounting basis:

1. Start with net income and add back relevant noncash expenses such as depreciation.
2. Reconcile changes in receivables, inventory, contract assets, payables and other operating balances under the approved accounting policy. Avoid counting unbilled work twice.
3. Separately record capital purchases, debt principal, owner distributions and new financing.
4. Add the resulting change in cash to opening cash and match ending bank cash.
5. Investigate unsupported source links, cutoff differences and classifications before deciding what caused a shortfall.

Use the cash gap calculator for a planning scenario. A calculated scenario does not establish the company's available bank balance.

Why the two numbers disagree

On the assumed accrual basis, revenue follows the approved recognition policy while bank cash follows actual receipts and payments. An invoice or a field completion status alone does not establish either.

The four checks below help locate a timing, process or source gap. They may also explain a legitimate difference. Availability depends on the records and history in the installed systems.

Gap 1: Logged Hours Versus Invoiced Labor

Across 963 companies with at least 100 logged job hours, the median invoiced-to-logged job-hour ratio is 97.1%, P25 89.2% and P10 66.9%. The denominator includes only hours assigned to jobs, not all paid technician hours.

A lower ratio is a reason to check billed-hour units, period alignment, job matching, flat-rate or fixed-price work, warranty/callback hours and logging completeness. Some recorded ratios exceed 100%; their causes are not measured by this distribution. Candidate missed billing needs a contract-eligible job-to-invoice trace before anyone calls it a loss.

If hours were worked but never logged, this ratio cannot see them. Keep payroll-to-time completeness and logged-time-to-invoice matching as separate checks.

Gap 2: the invoice that sat on someone's desk

Across 733 companies the median time from job completion to invoice issued is 1 day, which sounds like a solved problem until you look at the distribution. Among companies that invoice after completion rather than progress-billing, the median is 7 days. The published bottom-decile value is 30 days, but the canonical summary does not identify whether that decile covers all 733 companies or only post-completion invoicers.

If payment terms run from invoice date, a 30-day invoicing lag on Net 30 terms can put payment roughly 60 days after completion. Check the contract and receipts before inferring customer financing.

Gap 3: Billed Revenue Versus Recorded Collections

Across 464 contractors, recorded cash collected / billed revenue has median 85.1%, P75 92.7%, P25 70.7% and P10 38.8%. The comparison window is not published.

For illustration, applying 85.1% and 96% to the same fictional $5M eligible billed population gives $4.255M and $4.8M of cash, a $545K difference (about $550K). This is arithmetic under invented comparable-population assumptions, not an observed company's AR, permanent noncollection or a forecast of recoverable cash.

Read the actual invoice-aging, retainage, credit, write-off, unapplied receipt and cutoff bridge. Billed revenue and recognized profit are different measures; the change in supported receivables belongs in the cash bridge.

Gap 4: the job that was finished but never closed

The canonical closeout metric has n=555 completed-job observations: median 1.7 days, P75 9.6 and P90 35.8 from work complete to recorded financial job closeout. Lower is faster; P75 is the slower side of this interval.

This status interval does not establish invoice issuance, retainage release or payment timing. Inspect those records separately. An open financial file can contain paid invoices or legitimate pending cost approvals; the closeout marker is a review cue, not proof that cash is trapped.

What this looks like together

Keep these fictional comparisons separate:

• Profit: $4M recognized revenue at an assumed 12% net margin is $480K net income. Only a balance-sheet and cash-flow bridge explains the change in bank cash.
• Hours: On an invented 40,000 logged-hour population, 97.1% versus 89.2% corresponds to 38,840 versus 35,680 invoiced hours, a 3,160-hour difference. At an assumed $40 labor cost, that is $126,400 (about $126K) of cost-equivalent hours. It is not eligible selling value or collected cash, and may include legitimate nonbillable work.
• Collection: On the same fictional $4M eligible billed base and a comparable window, 85.1% versus 70.7% corresponds to $3.404M versus $2.828M cash, a $576K modeled difference. It is not a point-in-time receivable balance or an observed improvement.
• Timing: Compare actual invoice timestamps and financial-closeout timestamps separately. The published medians come from different samples and are not a single company's path.

Do not add the cost-equivalent hours to the cash-ratio difference and call roughly $700K earned cash. Nor can separate percentile inputs explain the assumed $480K net income. Match the company's own period, eligible transactions and accounting treatment.

The diagnostic, in order

1. Billing capture: inspect invoiced / logged job hours for a defined period. A ratio under 90% is a review trigger, not proof of the largest leak. Check nonbillable scope and units first.
2. Invoicing lag: trace work complete to invoice sent, keeping progress-billed and post-completion work separate. Inspect terms and receipts before inferring customer financing.
3. Collection: reconcile eligible cash receipts and billed revenue over a stated window, including earlier invoices and cutoff differences. A 90-day local view cannot be directly declared above or below the published ratio whose window is unknown.
4. Closeout: review the status interval and the supporting invoice, cost and approval chain. More than ten days open does not establish that cash is uncollected.

Use supported records or exports and preserve original IDs. If your field software and accounting disagree on those same records, see Why Your Field Software Numbers Don't Match Your Accounting.

One number to watch that is not about cash

Across 959 companies with at least $100K of revenue, the median gets 31% of revenue from its single largest customer, and a quarter of companies are above 54.6%. Concentration does not cause a cash gap by itself. It decides how bad the gap gets when one customer slows down, because at 55% concentration a single late payer is not a collections problem, it is a much larger cash risk.

FAQ

Why is my company profitable on paper but always short on cash?

Profit, recognized revenue and bank cash follow different events. Reconcile opening cash plus operating, investing and financing movements to ending cash. Receivable growth can be one explanation, but the collection-ratio distribution does not measure your receivables or permanent loss.

What is a normal collection rate for a contractor?

The cross-trade observed median is 85.1%, P75 92.7% and P10 38.8% across 464 contractors. Its measurement window is unpublished, so match the eligible invoice/receipt population and timing before making a local comparison.

How many of my logged hours should I be invoicing?

Median recorded invoiced / logged job hours is 97.1%, P25 89.2%, P10 66.9% across 963 companies. Job-assigned hours are the denominator. Check units, warranty work, fixed-price billing and logging completeness before defining an appropriate local target.

Which of the four gaps should I fix first?

Start with a reproducible record population and the largest supported exception affecting the current decision. Missing hours, incomplete invoice history and delayed receipts require different evidence; no benchmark makes one the universally largest issue.

Source and claim note: All figures here are from the Level Index, our own analysis of contractor financial and job data, with the sample size stated for each metric so you can weigh it. Sample sizes are 464 contractors for collection rate, 963 companies for billing capture, 733 companies for billing speed, 555 completed jobs for closeout lag, and 959 companies for customer concentration. Percentiles are the observed distribution, not a modeled one. Directional statements about what a specific number means for your business are interpretation, not measurement.