Why Your Field Software Numbers Don't Match Your Accounting

Your integration says it synced. What should you check?

Do not start with two dashboard totals. Pick one closed month and compare the same jobs in both systems:

1. Identity. Does each invoice, payment, credit, job, customer, and cost record have a matching source record?
2. Balance. Do original amount, payment amount, credit, write-off, and remaining open balance agree?
3. Timing. Are you comparing the same operational date, accounting posting date, and report cutoff?
4. Definition. Are both reports using the same job status, cost basis, revenue treatment, and included record population?

If one record cannot be traced end to end, the report total is not yet a decision-grade number. Download the worksheet below and classify the difference before rebuilding the integration or changing the books.

Check whether the two systems are answering different questions

Your field software knows what a technician did and what was quoted. Your accounting system knows what was invoiced, what was paid, and what hit the general ledger. Those overlap, and they are not the same thing.

One system may contain an error, or both may be correct on different definitions. Test the three differences below before deciding whether the gap needs a correction or an explanation.

Difference 1: manual handoffs can leave records out of sync

JBKnowledge's 2021 Construction Technology Report reported 51% transferring data between applications manually. This describes that survey, not today's contractor population. Manual handoffs also occur in Level CFO's field work; those observations are not a representative survey.

Manual transfer does not just risk typos. It means the two systems are only as synchronized as the last time someone did the export, so month-end comparisons are usually comparing two different moments.

Named studies report different field-service software adoption results by trade, but they are not a comparable benchmark set:

Read that table as a source register, not a cross-trade ranking. The surveys use different methods and years, so their percentages are not a clean per-trade comparison. Do not use the rows to rank adoption, calculate an average, or infer a current market share. The useful operating question is simpler: reconcile one month of field hours to invoiced hours, whether your workflow is integrated, exported, or entered by hand.

Difference 2: the hours that never became invoices

This is the one that shows up as a margin mystery, and it is measurable in your own data today.

Across 963 companies, the median invoices 97.1% of the hours logged against jobs. The bottom quartile invoices 89.2%, and the bottom decile 66.9%.

Those hours have not appeared on invoices, but that does not mean they are absent from accounting. Payroll costs may already be recorded, and applicable accrual policies may recognize earned but unbilled revenue. Check whether each gap is delayed billing, included in a fixed-price contract, nonbillable work, or a missing charge before treating it as recoverable revenue.

State the denominator whenever you use this figure: it is hours already assigned to jobs. Hours that never made it onto a job at all are not in the 97.1%, which is why that number is the optimistic view of billing discipline rather than the flattering one.

Difference 3: the two systems recognize the same work at different moments

Job completion, invoicing, revenue recognition, and cash collection are separate events. Under accrual accounting, issuing an invoice is not a universal trigger for recognizing revenue. Recognition depends on the applicable policy and contract performance; cash-basis records follow a different basis. FASB's Topic 606 implementation guidance explains why a billing schedule alone does not determine recognition. Reconcile progress billing, retainage, change orders, and any WIP or unbilled balances with the person responsible for the books.

Across 555 completed jobs the median closeout takes 1.7 days, but the top-quartile boundary is 9.6 days. A closeout delay that crosses month end can move invoicing into another period. Whether recognized revenue also moves depends on the accounting policy and any closing adjustments.

What a real reconciliation looks like

Do it on one month, not the whole year, and in this order:

Download the mismatch worksheet

The worksheet is ungated. It preserves source and accounting record IDs, dates, amounts, statuses, dimensions, likely cause, owner, next action, and remaining uncertainty.

1. Hours. Total hours logged to jobs in field software against total hours invoiced in accounting. The difference is your billing capture gap, in hours, before anyone argues about dollars.
2. Revenue. Field software revenue for jobs completed in the month against accounting revenue recognized in the month. Expect a difference. You are looking for whether it is explained by billing timing, recognition adjustments, or missing records.
3. Cost. Field software job cost against the GL cost for the same jobs. This is where subcontractor invoices arriving late usually surface.
4. Margin. Only now compare margin. Median job gross margin is 44.3% across 1,747,089 completed jobs, with the bottom-quartile boundary at 28.4%. Use those as cohort context, not a diagnostic threshold. Investigate missing costs by matching actual cost records for the same jobs, not by comparing your margin spread with a different population.

The output should be a documented reconciliation with explained differences and unresolved exceptions. Use it to decide whether you need an integration correction, a billing process, or closeout discipline. The Financial Systems Field Guide covers system ownership, mappings, and the checks to carry into a setup assessment.

The one that is worth fixing first

Billing capture is a useful first diagnostic when job-hour and invoice exports are available. In our experience, reconciling one month first is usually faster and less disruptive than starting with a full integration project because it identifies whether the issue is a billing process, a record-mapping failure, or a system gap. Some unmatched records may support a valid late invoice; others may already be included in a fixed-price contract or be nonbillable. Integration work can help recover missing records and prevent recurrence. Choose the next step from the size and cause of the verified gap.

Get a two-system mismatch review

Level reviews the records, mappings, cutoffs, and exception ownership behind your field and accounting systems. The first output is a prioritized reconciliation plan that identifies what differs, why it differs, who owns the next action, and what still needs evidence. It is not a generic software replacement recommendation.

Request a two-system mismatch review

FAQ

Why do my field software numbers not match my accounting?

Manual handoffs, different timing, and different definitions can explain a gap, but errors are also possible. Compare the same jobs and period, then reconcile invoices, recognized revenue, payroll, and other costs. Unbilled work may already appear in payroll costs or applicable accrual balances.

What share of contractors use field service management software?

Two dated studies report results for roofing and electrical contractors, but their methods, populations, and years differ. Use them for context, not as a current cross-trade ranking or a Level benchmark. Your own reconciliation result is more actionable than a broad adoption percentage.

Do I need an integration to fix this?

First identify the missing or mismatched records in a bounded sample. A billing-process fix may be enough; an integration correction may be needed if records are missing or mapped incorrectly. Neither the time required nor the recoverable amount is known until the exceptions are checked.

What is a normal job gross margin to expect in field software?

Median job gross margin is 44.3% across 1,747,089 completed jobs, with the bottom quartile at 28.4%. Those describe the measured cohort, not a universal target or expected spread for your business. Match job type, cost definitions, and recognition timing before comparing your margins.

Source and claim note: Billing capture, job gross margin, and job closeout are from the Level Index, our own analysis of contractor financial and job data. Sample sizes are 963 companies for billing capture, 1,747,089 completed jobs for job gross margin, and 555 completed jobs for closeout lag. The manual-transfer figure is from JBKnowledge's 2021 Construction Technology Report; disconnected systems are also a Level field observation. The trade adoption figures are preserved with their named studies as a source register, not a comparable or current cross-trade distribution. Do not use them as a Level benchmark.