Home › Guides › Costing & Finance
Guides

Restaurant cash count: before vs after with Masterestaurant

Diego F. Parra By Diego F. Parra · Updated · Costing & Finance
Restaurant cash count: before vs after with Masterestaurant — Masterestaurant
Quick verdict

Verdict: a restaurant cash count works when it is done blind, by shift and against the POS close, with a fixed float and a tolerance written down before anyone counts; without that protocol, counting bills at midnight only confirms what the cashier already knew you would find.

And margins leave no room: the National Restaurant Association puts the median pre-tax profit of US full-service restaurants at 2.8% of sales (2024), so a shortage nobody investigates eats several days of profit.

🧭 GuideStep-by-step guide with a measurable outcome per step· 14 min read· 2026-10-28

A restaurant cash count is not counting the money at close. It is proving that what sits in the drawer matches what the system says should be there, shift by shift, and explaining every difference before the team goes home. That distinction looks minor, and it is exactly what separates an operation that controls its cash flow from one that finds the holes when the bank no longer matches the P&L.

Cash also weighs more than many owners assume. In Spain, 57% of in-store transactions were paid in cash in 2024, according to the Banco de España; in Mexico, INEGI and CNBV's Enafin 2024 shows cash as the payment method customers use most in 79.8% of business cases. Dropping the count because almost everything is cards means you stop watching a BIG share of sales at the point where money changes hands most often.

Before your first count with the method you need four prerequisites: a fixed float per drawer, the POS close report by shift and cashier, a count sheet by denomination, and a written tolerance that says from which difference you investigate. Diego F. Parra sets an order Masterestaurant does not negotiate: count first, look at the system second, because whoever knows the expected figure tends to find the missing money.

Side-by-side comparison

Side-by-side: restaurant cash count

Before: eyeball countAfter: Masterestaurant method
When you count✕Once, at the end of the night, if the manager can stay✓At the end of every shift and every cashier change, at a fixed time
Who counts and how✕The cashier who rang the sales, with the POS report in view✓Blind count by the cashier; a second person checks against the POS afterward
Drawer float✕Variable: whatever was left yesterday✓Identical fixed float every opening, pulled out before counting sales
Variance tolerance✕None; decided on the spot✓Written per shift before counting; anything above it is investigated the same day
Overages✕Kept, or treated as good news✓Logged and investigated exactly like a shortage
Payment types✕Only cash is checked✓Cash, card terminal, transfers and delivery apps reconciled separately against the POS

What fixed float does each register need before the first cash count?

Each register needs an identical fixed float at every opening, with its breakdown by denomination written down, because without a constant starting point no difference at close can be explained.

You set the amount from your average ticket and the change that shift usually requires, and it is handed over counted and signed by the person opening and the person receiving. For example, if your average ticket is around 25 and most guests pay with large bills, the float carries enough coins and small bills so nobody leaves to find change mid-rush, exactly when the drawer sits open and unwatched.

What is done is a float sheet by denomination, with two signatures and the time, filed with the shift report. If the float changes from one day to the next, the count loses its reference and you are counting blind in the WORST sense.

The blind count: drawer first, system second

Counting blind means the person counting the cash does not see the expected POS figure until they have signed their own total, and it is the rule that exposes the most honest differences. At the end of the shift, the cashier counts by denomination on the count sheet, sets the fixed float aside, records card vouchers and partial drops, signs and hands over; only then does the manager print that cashier's close.

Diego F. Parra holds to this order within the Masterestaurant method because anyone who knows the target number tends to recount until it appears, or stops counting the moment it balances. And the volume justifies the discipline: the European Central Bank (SPACE, 2024) found that 52 % of point-of-sale payments in the euro area were made in cash. Verifiable deliverable: the signed sheet, time-stamped before the POS report was printed, something any auditor can confirm in ten seconds.

Which POS figure is the cash count compared against?

The count is compared against the POS close for that shift and that cashier, split by payment method, never against total sales for the day.

The expected-cash formula is simple to write and easy to forget: sales collected in cash plus the fixed float, minus partial drops to the safe, minus petty-cash payouts with their receipt. Cash tips go on a separate line, because mixed into sales they create false overages that cover real shortages.

In markets where cash still rules, this line IS the sales figure: in Colombia, the 2023 Banco de la República survey put cash at 78.6 % of the value of everyday payments, according to Portafolio (2024). This step produces a reconciliation sheet with three lines (expected, counted and difference) signed by cashier and manager, with a difference you can trace to a ticket or a drop.

Written tolerance and what to do with every difference

Tolerance is set in writing before counting, as a concrete amount per shift, and any difference above it is investigated that same night, whether short or over. An overage deserves the same scrutiny as a shortage, because it almost always hides an unrecorded sale or a tip that slipped into sales.

The margin explains the strictness: the National Restaurant Association (2025) puts the 2024 median pre-tax income at 4.0 % of sales for limited-service and 2.8 % for full-service restaurants in the U.S. For example, in a restaurant selling 10,000 a day, a daily shortage of 100 equals one full point of sales, about a third of that full-service margin.

It is done when a variance log records cause and owner, and none rolls into the next day without a written explanation.

Count every shift, even when little cash comes in

Count every shift even if lunch is paid almost entirely by card, because skipping one contaminates the next. Think through what happens if you drop it: the night shift inherits a float nobody verified, any evening difference can no longer be separated from the lunch one, and weeks later the accumulated gap shows up at month-end as inflated food cost, which you will blame on the kitchen when the problem was in the drawer.

The bet that cash is disappearing also fails against small-purchase data: according to the ENIF 2024 by INEGI and CNBV, 85 out of 100 Mexican adults prefer paying cash for purchases of 500 pesos or less. The verifiable output is a count calendar with no gaps, a sheet and signature per shift, and one written rule: the float never changes hands uncounted.

The mistakes that ruin a well-designed cash count

The costliest mistake is letting the cashier see the POS report before counting, because it turns the count into a balancing ritual instead of a control. Close behind come a float adjusted "by eye", drops to the safe with no receipt, tips in the sales drawer and a tolerance that exists only verbally. I got this wrong for years: I believed counting at the end of the night was enough, and got a figure impossible to attribute to anyone.

The paradox is that the protocol protects the honest cashier most of all, because their difference stays in their shift, with their signature, and does not turn into suspicion over the whole floor. And do not wait for cash to vanish: the European Central Bank (SPACE, 2024) counted 88 % of euro-area businesses still accepting it that year, compared with 96 % in 2021.

How do you know the cash count was done right?

The cash count was done right when every shift has a signed float, a blind count time-stamped before the POS report, a reconciliation by payment method and every difference above tolerance explained in writing.

To close the week, check five things: the float was the same at every opening, no shift was left without a sheet, tips traveled on their own line, drops to the safe carry a receipt, and overages were investigated as rigorously as shortages. If one fails, fix it before month-end close, when the difference has already blended into the income statement.

And cash still carries weight in Europe: 39 % of the value paid at the point of sale in 2024, according to the European Central Bank. Your action for tomorrow: print the count sheet and test it on the first shift.

What really changes between before and after?

The paradox of the cash count is that the more you trust your team, the more you need the protocol. A blind count protects the honest cashier:

the variance is documented on their shift, with their signature, instead of turning into vague suspicion of everyone, which is what actually breaks trust on a floor. What happens if you stop counting the lunch shift because hardly any cash comes in?

Dinner inherits a float nobody verified, night variances can no longer be separated from lunch, and a few weeks later the accumulated shortage shows up at month end disguised as inflated food cost or a drop in contribution margin you will blame on the kitchen.

What really changes between before and after — in practice?

In Diego F. Parra's method the count does not live alone at the register:

it feeds the weekly Masterestaurant dashboard next to theoretical vs actual food cost and break-even, because a persistent cash variance and inventory shrink that won't reconcile usually share one origin. According to the European Central Bank's SPACE study (2024), cash still accounted for 52% of point-of-sale payments in the euro area, so the drawer is not going away.

Point by point

A/B analysis: eyeball count vs method count

Catching shortages
A · Before: eyeball countFound at month end, when nobody remembers the shift
B · MasterestaurantFound on the shift where they happen, with signature and time
Verdict: After wins: a same-day shortage gets explained, a month-old one gets paid.
Reading costs
A · Before: eyeball countCash variances get confused with high food cost or shrink
B · MasterestaurantWith a clean drawer, theoretical vs actual cost reads without noise
Verdict: After wins: without a balanced drawer you cannot honestly tell whether a dish respects the method's 32% food cost ceiling, which is a maximum.
Closing time
A · Before: eyeball countA few minutes
B · MasterestaurantSomewhat longer at first, then nearly the same
Verdict: Honest concession: before is faster. Those saved minutes are paid for with money you never see.
Sizing cash
A · Before: eyeball countCash is assumed to be marginal
B · MasterestaurantSized with market data: in Colombia cash made up 78.6% of the value of everyday payments, per the Banco de la República survey reported by Portafolio (2024)
Verdict: After wins: the count is designed for the cash that actually comes in.
Side-by-side comparison

Eyeball count (before)

  • Counted at midnight, in a hurry.
  • The cashier sees the expected POS total first, so any small difference gets fixed with tips or the float and the owner never learns it existed.
  • Variable float.
  • No signed sheet, so nobody can tell which shift a hole started in.

Counted with the Masterestaurant method (after)

  • Blind count per shift.
  • A fixed float pulled out before counting sales, always the same amount, so cash sales come out clean and comparable from one day to the next.
  • Written tolerance.
  • Two signatures, and each shift's variance goes into the weekly log management reviews next to prime cost.
The numbers that matter

The cash your count has to reconcile, in numbers

52%
of point-of-sale payments in the euro area were made in cash in 2024 (59% in 2022)
39%
of the value of euro area point-of-sale payments was cash in 2024 (42% in 2022)
57%
of in-store transactions in Spain were paid in cash in 2024
79.8%
of cases: cash is the payment method customers of Mexican businesses use most
85.2%
of Mexico's population prefers cash for purchases up to 500 pesos
78.6%
of the value of everyday payments in Colombia is made in cash (74.6% in the previous survey)
4.0%
of sales: median pre-tax profit of US limited-service restaurants, 2024
2.8%
of sales: median pre-tax profit of US full-service restaurants, 2024
88%
Share of euro area businesses accepting cash in 2024 versus 2021 (context for how much cash businesses still handle and must reconcile in a cash count)
Visualization
The numbers, visualized
The numbers, visualized52% of point-of-sale payments in the euro area were made in cash; 39% of the value of euro area point-of-sale payments was cash in; 57% of in-store transactions in Spain were paid in cash in 2024; 79.8% of cases: cash is the payment method customers of Mexican bu; 85.2% of Mexico's population prefers cash for purchases up to 500 ; 78.6% of the value of everyday payments in Colombia is made in casof point-of-sale payments in the euro area were made in cash in 2024 (59% in 2022)52%of the value of euro area point-of-sale payments was cash in 2024 (42% in 2022)39%of in-store transactions in Spain were paid in cash in 202457%of cases: cash is the payment method customers of Mexican businesses use most79.8%of Mexico's population prefers cash for purchases up to 500 pesos85.2%of the value of everyday payments in Colombia is made in cash (74.6% in the previous survey)78.6%
Sources: European Central Bank, SPACE study 2024 · Banco de España, Economic Bulletin 2025/Q2 (via Bolsamanía, 2025) · INEGI and CNBV, Enafin 2024 (cited in El Financiero, 2025) · Xataka México citing INEGI, ENIF 2024 (2025) · Portafolio with Banco de la República data (2024)Chart by masterestaurant.com
Illustrative case (composite)

“For four weeks the night close always matched to the cent, and that was exactly what felt wrong; once we moved to blind counts per shift and pulled the fixed float first, overages showed up in three of fourteen shifts in the first week, all counter sales with no ticket, and we finally saw where our margin was going.”

— owner of a two-location ceviche restaurant in Guayaquil, illustrative case

Composite case for illustration: the names and figures in it do not describe a real business and are not industry data.

How to apply it in your restaurant

How to run a restaurant cash count in 4 steps

Set the fixed float and pull it before counting
Deliverable: every drawer opens with the same float, in set denominations, recorded on the count sheet; at close, that exact float comes out first and cash sales are counted separately. Checkpoint: the float pulled at close equals the opening float, zero difference. Typical mistake: topping up the float with sales cash, the fastest way to erase the trail of a shortage.
Count blind, by denomination
Deliverable: a sheet signed by the cashier with the count by denomination, done without seeing the POS expected total. Only then does a second person print the shift close and compare. Checkpoint: two independent counts of the same drawer match; if not, count a third time before looking at the system. Typical mistake: handing over the POS report with the blank sheet.
Reconcile every payment type against the POS
Deliverable: one line per payment type (cash, card terminal, transfers, delivery apps) with expected, counted and variance. Checkpoint: the sum of all types equals total shift sales in the POS, and each variance sits inside the written tolerance or carries an explanation. Typical mistake: offsetting a cash shortage with a card overage; those are two errors, not one that cancels out.
Log, investigate and bring the variance to the dashboard
Deliverable: a weekly variance log by shift and cashier that management reviews with prime cost and cash flow. Checkpoint: no out-of-tolerance variance goes without a written cause by the next day. For example, if one drawer piles up several small variances in the same direction over a month, treat it as a pattern even if each one alone was within tolerance. The math is simple: the National Restaurant Association puts median pre-tax profit for US limited-service restaurants at 4.0% of sales (2024), so a small repeated hole weighs far more than it looks.
✦ AI applied

And with AI?

Project your food cost, spot margin leaks and simulate pricing scenarios in minutes. Diego F. Parra is an expert in AI applied to restaurants.

Masterestaurant tools & method

Method tools to close the drawer with control

The cash count is the first line of financial control, not the last: a drawer that balances every day only tells you money wasn't lost between the guest and the register. You still need to know whether each dish leaves margin and whether sales cover break-even, and these Masterestaurant tools connect the close to those questions.

Diego F. Parra

Diego F. Parra — International consultant, expert in creating and scaling restaurants and in AI applied to restaurants, foodtech and HORECA. Methodology applied in 8.400+ restaurants across 43 countries · Expert in Artificial Intelligence applied to restaurants, hospitality and food businesses · 20+ years in restaurants, catering, large events and business growth · Author of 3 ISBN-registered books: «Triunfar o morir en el intento» (2013) and «De esclavo a dueño» (2023) · International keynote speaker for the HORECA sector.

FAQ

Restaurant cash count FAQ

What is a restaurant cash count?

It is counting the money in each drawer at the end of the shift and comparing it with the sales the POS recorded by payment type. The goal is to find and explain variances the same day, not just to know how much cash came in.

What is a restaurant cash count?

It is counting the money in each drawer at the end of the shift and comparing it with the sales the POS recorded by payment type. The goal is to find and explain variances the same day, not just to know how much cash came in.

How often should a restaurant do a cash count?

At the end of every shift and at every cashier change, at minimum. A single daily count mixes shifts and people, so when a shortage appears there is no way to know where it started or who should explain it.

How often should a restaurant do a cash count?

At the end of every shift and at every cashier change, at minimum. A single daily count mixes shifts and people, so when a shortage appears there is no way to know where it started or who should explain it.

What cash variance is acceptable in a restaurant?

The one you write down before counting, never one decided on the spot. Anything above that tolerance is investigated the same day, and overages get the same treatment as shortages.

What cash variance is acceptable in a restaurant?

The one you write down before counting, never one decided on the spot. Anything above that tolerance is investigated the same day, and overages get the same treatment as shortages.

Is a cash count worth it if almost everything is paid by card?

Yes, because cash is still there: the ECB's SPACE study found 88% of euro area businesses accepted cash in 2024. A method-based count also reconciles card terminals, transfers and apps against the POS.

Is a cash count worth it if almost everything is paid by card?

Yes, because cash is still there: the ECB's SPACE study found 88% of euro area businesses accepted cash in 2024. A method-based count also reconciles card terminals, transfers and apps against the POS.

Data & sources

2026 data on restaurant cash count

Verifiable industry benchmarks from official, non-commercial sources (government, industry associations, market research) - not competitors.

MetricValueSource
SBA 504 loans approved in fiscal year 2025 in the U.S., relevant for restaurant real estate or equipment purchasesUSD 7,8 mil millones en 6.750 préstamos 504, FY2025U.S. Small Business Administration — Trump SBA Delivers Record Capital to Small Businesses in FY25 (2025)
Total SBA 7(a) and 504 loans guaranteed in fiscal year 2025 in the U.S. (SBA loans for restaurants, all industries)USD 44,8 mil millones en 84.400 préstamos, FY2025U.S. Small Business Administration — Trump SBA Delivers Record Capital to Small Businesses in FY25 (2025)
SBA loans to small business startups in the U.S. since January 2025, relevant for SBA loans for new restaurants8.900 préstamos a startups por USD 5,6 mil millonesU.S. Small Business Administration — Trump SBA Delivers Record Capital to Small Businesses in FY25 (2025)
Maximum amount of an SBA 7(a) loan for restaurants and other small businesses in the U.S.USD 5 millonesU.S. Small Business Administration — 7(a) loans (2026)
Maximum spread over the base rate on SBA 7(a) loans of $350,000 or more for restaurantstasa base + 3,0 %U.S. Small Business Administration — 7(a) loans (2026)
Maximum spread over the base rate on SBA 7(a) loans of $50,000 or less (small startup financing for a restaurant)tasa base + 6,5 %U.S. Small Business Administration — 7(a) loans (2026)

Restaurant cash count: the Masterestaurant method

Applied in +8.400 restaurants across 43 countries.

Community

Join our MASTERESTAURANT Community for FREE

Restaurant owners and teams from 43 countries sharing knowledge, tools and applied AI — straight to your WhatsApp.

Join the community
Author: Diego F. Parra  ·  Publisher: MASTERESTAURANT®
Content created with AI assistance, reviewed by the MASTERESTAURANT editorial team.
MR Comparison Engine v0.9.394