What four restaurant cash reviews found inside labor, food and void reports

A restaurant's void report included a dessert priced at $250.

The review classified it as a banquet package placeholder based on its pricing. Room fees and deposits were mixed into the same report. Across January through August 2026, banquet bookkeeping accounted for 52% of recorded void dollars.

An owner looking only at the total could walk into a meeting ready to confront the floor staff. First, someone needed to explain what those transactions represented.

That was one finding in four restaurant cash recovery plans I reviewed. Others included food invoices missing from a reporting system, recurring overtime, blank discount explanations and cash records that still needed to be matched to the bank.

The plans contain findings and proposed fixes. They do not establish money recovered after those fixes. I have kept the restaurants anonymous and the figures tied to the periods reviewed.

The food-cost report left out a major supplier

At a po'boy restaurant, the review found 54 vendor invoices totaling $107,721 from July 6 through September 2, 2026. None appeared in the reporting system's invoice or purchase records.

In July alone, that system recorded $5,219 in purchases across all vendors. The missing supplier's invoices totaled $54,139 for July.

The system was also showing zero income and zero labor.

I would not ask the kitchen to defend a food-cost percentage built from those records. The first job was to get revenue, labor and the missing invoices into the report, then verify a complete month.

The plan assigned no recovered dollars to this finding. The invoices documented food purchases that were absent from the reporting records. They did not show a refund due or food that had gone missing.

For an owner, the practical check is small enough to begin before the next cost meeting. Pick a major supplier, pull its invoices for the reporting period and trace them into the report. Include credits. Check that the dates and amounts agree. If they do not, resolve that difference before using the percentage to change purchasing or portions.

Once the report includes the purchases, you can investigate the cost itself. A separate restaurant's January-July 2026 invoice review flagged per-unit price changes on flounder, rockfish, halibut and salmon.

Before disputing an invoice, compare the same product specification, pack size and unit, delivery date, and agreed or reference price. The review had not established an overcharge.

Closing labor needs a shift attached to it

At the restaurant with the banquet business, closing overruns were being checked through camera footage.

The review matched 7,327 scheduled shifts to actual time records across January-August 2026. It found 1,671 unscheduled hours, concentrated in cook and dish roles on Fridays and Saturdays. Twenty-nine percent of matched shifts ran more than 15 minutes past scheduled close.

The calculation applied a 15-minute buffer and a three-hour outlier ceiling. The ceiling was intended to limit the effect of possible forgotten clock-outs.

The records gave the manager a specific place to look. They did not establish that every extra hour was avoidable.

I'd want the manager to review a flagged closing shift while it was still fresh. Check when the last orders arrived, what work remained and whether the schedule allowed enough time to finish it. Use that record to decide whether the close needs a longer schedule, earlier cleanup or a corrected clock entry. Test the change on the same role and day of the week with a comparable workload; check paid hours and completed closing work together.

The plan proposed an alert 15 minutes past scheduled close and a manager to review it. At the time of the plan, that work was still pending.

Another restaurant group had a cook-labelled record with overtime in every month from January through July 2026. The review counted 837.89 overtime hours under the matching record label.

Employee IDs were blank. Finance still needed to confirm the person, paid cost and cause. The pattern warranted a conversation, but the report could not tell the owner to remove those hours.

Before moving a shift to someone else, confirm that person can do the work and what the replacement hours will cost. Check that the change does not move the overtime to another employee. Then compare total paid cost and whether the work got done. A smaller overtime line is only part of that check.

Sometimes the schedule itself is the missing evidence. At a mixed-operation location in a burger group, the review found 3,288 scheduled hours against 8,647 worked hours for January-July 2026. On the review's basis, the schedule covered 38.0% of worked hours; it excluded 404 unclocked scheduled hours from that comparison.

The source explicitly treated this as a schedule-coverage finding. Calling the rest overspending would assume the schedule was complete enough to serve as a budget. The next step was to establish which work and roles it actually covered.

A blank discount reason does not tell you which policy to change

The burger group's January-July 2026 discount export contained 18,573 rows totaling $166,867. Every reason and comment field was blank.

The review traced 57.7% of discounted dollars to loyalty-system activity, 37.1% to promotion buttons and 5.2% to individual discretion.

Those categories call for different checks. Review the loyalty rules and who can change them. Check which promotion buttons should still be available and under what conditions. Require a reason and an approval record for discretionary discounts.

Filling in a reason field would improve the record. It would not, by itself, determine whether a promotion was worth running.

A separate review at the po'boy restaurant showed why suspicion needs to stay separate from the finding. Its May 26-July 31, 2026 discount records also had blank reasons on every row. Yet 96.8% of discounted dollars came from named programmatic promotions. The plan claimed no recovery on that finding.

The owner still had a documentation problem to fix. But the report gave them a reason to examine promotion rules before accusing staff of handing out unauthorized discounts.

The cash overage still needed a bank statement

At the same restaurant, the POS cash summary showed an overage of $449.61 in June and $5,748.39 in July 2026. The safe's check kept showing $0.00, including a run of 20 straight days.

No bank statements were included in that review. It left the question open and claimed no dollar loss.

An overage is a difference to explain. I'd ask the owner and bookkeeper to match the same dates across the POS, safe records and bank, accounting for payouts, tips and deposit timing. I'd also ask what the safe's zero measured and whether the check was running as intended.

The reconciliation should show expected cash, logged safe and deposit amounts, bank credits, and the entry that explains each difference. Keep timing differences open until the corresponding credit appears.

Put the next check on the calendar

Each finding needs a record of what happens after the review. For closing labor, name the manager reviewing flagged shifts and the date for checking paid hours after any change. For missing invoices, have the bookkeeper verify the next complete month. For cash, keep the reconciliation open until the supporting entries explain the difference.

Start with the next flagged kitchen close. Have the manager bring the scheduled end, actual clock-out, last order time and unfinished work to the next shift review. Decide what to change, then assign that manager to check paid hours and completed work on the next comparable close.

Have Marty take the first pass

Marty is your outsourced Chief Profit Officer for Restaurants. Hire Marty to do the records work behind the cost problem you want solved. We start with a paid Cash & Profit Review and agree on the records, access and fee before work begins.

We examine the transactions, flag what still needs an explanation and recommend who should do what next. You get a plan your managers and bookkeeper can work from, with the supporting records beside each finding.

Your team can carry out the plan. Ongoing help is optional and agreed separately.

Start Your Review: https://usemarty.com/demo

Source note: This article draws on four anonymized Marty cash recovery plans supplied for review on September 9, 2026. The two September plans are dated September 4 and September 5, 2026; the burger-group source records a rerun on August 28, 2026; the other restaurant-group plan is undated. Observation periods appear beside the figures. The underlying exports were not independently audited for this article. None of these examples is presented as verified money recovered.