Chapter 34 — Further Reading

Sources are grouped by the book's three citation tiers. Tier 1 is verified canonical material we can stand behind. Tier 2 is real industry practice and benchmark ranges whose exact citation we have not pinned down — treat these as ranges, never as decimals to quote. Tier 3 is illustrative and constructed: Bellwether, both case studies, and every worked figure in this chapter.

One warning belongs at the top of this list rather than buried in it. This is the chapter of the book where the published literature is least trustworthy. Loss-prevention writing in the restaurant trade is substantially produced by people selling loss-prevention products, and it leans on a percentage-of-sales theft figure that cannot be sourced. Read the material below for structure — what a control is, what a report proves, what the law requires — and get your numbers from your own count sheets.

TIER 1 — Verified canonical

The COSO Internal Control—Integrated Framework. First published in 1992, updated in 2013, and the reference definition of internal control in American practice. Its components — control environment, risk assessment, control activities, information and communication, and monitoring — are where §34.1's authorize / record / custody / reconcile structure comes from. Read it for the definition and for one specific idea worth more to a small operator than any checklist: the control environment, meaning what the organization's leadership actually rewards and tolerates, is the foundation every other control sits on. That is §34.8's argument in an accountant's vocabulary.

The Sarbanes-Oxley Act of 2002, and internal control over financial reporting for public companies. You are almost certainly not subject to it. It appears here because it is why enterprise restaurant software has a mature exception-reporting module — and why, as Case Study 1 argues, an independent restaurant inherited one. Read the structure of the reporting obligation, not the compliance detail.

The Payment Card Industry Data Security Standard (PCI DSS), maintained by the PCI Security Standards Council. A condition of accepting cards, not an option, and the reason your point-of-sale system requires a unique identifier for every user and logs access to system components. Unique credentials are what make every exception report in §34.3 attributable to a person at all. Read your own processor's merchant guidance, and note that the standard has been revised through several major versions — check which one your processor holds you to.

The EMV liability shift, October 2015 (United States, most in-store transactions). Public record. It moved liability for certain counterfeit-card losses toward whichever party held the lesser technology and triggered the payment-hardware replacement cycle that carried modern software into a great many independent restaurants.

Electronic sales suppression as a tax-administration category. The OECD published work on "zappers" and phantom-ware in the early 2010s; a number of US states have enacted statutes criminalizing the sale, possession, or use of automated sales-suppression devices, beginning in the early 2010s; and Québec required restaurants to install a government-certified sales recording module, phased in around 2010–2011. This is the clearest documented case of a tax authority mandating a tamper-evident transaction log in restaurants specifically. Whether such a statute exists in your state is a question with a real answer — find it.

IRS Form 8027, Employer's Annual Information Return of Tip Income and Allocated Tips, required of large food or beverage establishments — generally those where tipping is customary and where more than ten employees worked on a typical business day. The IRS has also operated voluntary tip reporting agreement programs for the industry. Both rest on the establishment's own records, which is one more reason the point-of-sale system is a compliance artifact before it is a management tool. Tip rules change; verify current requirements with your accountant.

The Fair Labor Standards Act (FLSA) and your state's wage-and-hour statute — specifically the recordkeeping obligations and the treatment of unpaid work time. §34.1's Code and Compliance callout is the short version: a supervisor who can edit the punches of an employee they supervise is a control exposure and a much larger compliance exposure at the same time. Unpaid work time is generally recoverable as back wages, the remedy scheme commonly permits liquidated damages equal to the back wages, and the lookback extends from two years to three for willful violations, plus attorney's fees. State law is frequently more generous to the employee than federal law. Use an employment attorney for anything consequential.

The Association of Certified Fraud Examiners (ACFE) and its recurring research on occupational fraud. A real professional body publishing real research, and worth reading for its taxonomy — asset misappropriation, corruption, financial statement fraud, and the sub-categories underneath — which is more useful to an operator than any headline figure. Read the methodology section first: the sample is built from cases submitted by certified fraud examiners who investigated them, which makes it a sample of detected, investigated, reported fraud in organizations that employed a fraud examiner. That is a legitimate and informative sample. It is not a base rate for independent restaurants, and it cannot be converted into a percentage of your sales.

The FDA Food Code, for receiving. §34.7's twenty minutes at the delivery door is simultaneously a financial control and the point at which Chapter 25's temperature verification happens — cold holding at or below 41°F, with local adoption varying. The joint product is what makes the control's economics work.

Roger Fields, Restaurant Success by the Numbers. The most directly useful published treatment of restaurant cost control for an independent operator, and unusually honest about the gap between a system described and a system operated. Read his treatment of daily reporting alongside §34.4.

Brown and Rowe, The Restaurant Manager's Handbook. Encyclopedic and dry, and for once that is exactly what you want. Use it as a checklist when you write your own cash-handling procedure and your own receiving procedure — its coverage of the mechanical steps is more granular than this chapter's and will surface handoffs you forgot you had.

Danny Meyer, Setting the Table. Not a controls book, and included for one reason: §34.8's argument that a control policy has to be universal, routine, and explained rather than targeted and secretive is a hospitality argument before it is an accounting one, and this is the best available statement of the underlying case. Read it against §34.3's Hospitality callout, where a control policy that requires a manager's code for a nine-dollar dessert costs you the visit.

The Foodservice Management Professional (FMP) credential and its study materials. The cost-control and internal-control domains cover this chapter's territory in the vocabulary the industry actually uses, which is worth having when you are talking to a bookkeeper, a franchisor, or a lender's analyst.

TIER 2 — Attributed, specifics unverified

Employee theft as a share of restaurant sales. Handle this carefully, because it is the most frequently repeated unsourced number in the industry. Figures in the low single digits as a percentage of sales circulate widely — three to four percent is the most common — and we have found no rigorous source for any of them. Do not quote a percentage of sales as data, do not put one in a business plan, and treat any vendor or consultant who leads with one as having told you something about their evidence standards. The honest statement of what is known: employee theft in restaurants is real, it is more common in cash-heavy and beverage-heavy operations, published estimates vary enormously, and the only figure you can defend is one you measured in your own building. Bellwether's \$53,122 is measured, decomposed, and constructed — and not one of its six lines requires a dishonest employee to exist.

Acceptable inventory variance. Industry guidance commonly describes a well-run full-service kitchen as holding food variance within roughly one to two percent of theoretical usage, with bar variance held tighter because spirits are countable to the tenth. Ranges, not laws — and the useful version of the question is not "what is acceptable" but "what is normal for me," which requires three or four periods of your own history before it means anything.

Comp and discount benchmarks. Full-service guidance commonly puts comps at or below roughly one to two percent of gross sales, with wide variation by service style, check average, and how aggressively the operation uses service recovery. §34.3 uses 1.0% for Bellwether. There is no reliable public benchmark for void rates at all — voids depend on menu complexity, modifier design, and printer routing more than on anything else, so measure your own distribution and compare each employee to your house average rather than to a published figure.

Cash as a share of restaurant tender. Broadly reported to have fallen well below half of transactions and to have continued declining, with the pace accelerating after 2020 and wide variation by format, neighborhood, and daypart. §34.2 uses 9% for Bellwether as an illustrative figure. Get your own from a month of tender reports — it is a four-minute exercise and it will tell you how much of your control attention cash actually deserves.

Effective card-processing rates. Full service is commonly described in the range of roughly 2.3–3.2% of net sales depending on card mix, average ticket, pricing model, and negotiated markup. Chapter 26 established Bellwether's at 2.81%; §34.4's Friday runs 2.76% effective on a 2.6% + \$0.10 schedule. Compute your own as total fees divided by total card volume — it is almost never the headline rate on the contract.

Payroll burden rates. Employer payroll taxes, workers' compensation, and benefits are commonly estimated in the range of 10–20% of gross wages for a restaurant, with workers' compensation varying sharply by state and classification. This book uses 13.0% for Bellwether. It matters here because a control priced in labor hours — the weekly count, the second pair of eyes, the named receiver — is under-priced by roughly an eighth if you forget it.

Restaurant turnover. Widely described as running around 75% annually industry-wide, with enormous variation by position and market. It matters to this chapter for two reasons: rotation and mandatory absence are harder to run than they look when a third of the roster is under six months old, and Case Study 2's turnover cost is the line operators consistently fail to anticipate when an investigation goes badly.

Blind counting and surprise counting as standard practice. Both are widely recommended in industry guidance and inventory-software documentation, and both are widely skipped. The specific claim in §34.8 — that a surprise count recorded only when it finds something is a search rather than a control — is the book's own framing rather than a citable benchmark, but the underlying practice of recording every count is standard advice everywhere it is discussed.

TIER 3 — Illustrative and constructed

Everything about Bellwether is constructed for teaching. Specifically, from this chapter:

• **The exposure list, \$53,122**: food variance \$22,282 · the five bar leaks from Chapter 15 \$16,169 · comps above the 1.0% target \$8,680 · receiving error at 1% of \$334,776 of purchasing \$3,348 · cash loss at 1% of \$139,500 through the drawer \$1,395 · time-clock abuse on the employee side \$1,248. Stated as 3.4% of sales, 20% of the \$261,020 operating profit, and 28% of what remains after \$69,500 of debt service.
• **The control program, \$4,849**: jiggers and a pour-cost review \$233 · portion scale \$0 (owned) · weekly count at 2 people × 1.5 hours × \$22 blended × 52 weeks = \$3,432 · second pair of eyes on the nightly count \$1,184 (10 minutes a night at \$22 across 312 nights, plus \$40 of blind count sheets and a clipboard) · named receiver, standing delivery window, roughly \$1,530 unburdened. Recover half and \$4,849 becomes \$26,561, a 5.5× return.
• The control-function matrix in Figure 34.1 — five roles against four functions, showing that the chef-owner holds authorization and primary custody simultaneously and that there is no fifth manager to hand one to.
• The cash procedure: banks of \$250 on two drawers · drop at \$500 over bank · tolerance ±\$5.00 or 0.5% of cash sales, whichever is greater · review at three nights out of tolerance in four weeks or any single night over \$25 · cash at roughly 9% of sales, \$139,500 a year.
• The comp analysis: \$486 on \$31,196 of gross in the sample week (1.56% against a 1.0% target); the seven-service breakdown showing \$315 of the \$486 on Friday and Saturday dinner; Friday's five reason codes totalling \$187 across 12 items, of which \$76 bought back long tickets and \$16 carried no reason code at all; 34 comped items averaging \$14.29; and the tiered authorization ladder at \$15 / \$75 / above.
• The void distribution in Figure 34.3 — 41 voids for \$128 across eight service employees, with one outlier at 18 and four ordinary explanations before an extraordinary one.
• Figure 34.4, the Friday daily sales report — 143 covers, gross \$6,807.00 down to net sales of \$6,580.00, \$8,279.00 accounted for in two directions, \$972.00 expected in drawer against \$970.25 counted, and \$7,236.71 of predicted card settlement at 2.6% + \$0.10 across 118 transactions.
• The seven-service week reconciling to Chapter 31's \$30,400 of net sales, \$2,128 of tax, and \$32,528 collected.
• Figure 34.5, the Period 8 variance report — food sales \$86,400, ideal usage \$25,488 against actual \$27,202, six categories, +\$1,714 and +2.0 points, with meat and poultry carrying 62% of the variance on 39% of the ideal usage.
• The threshold table: category greater-of 5% or \$250 · liquor and wine greater-of 3% or \$150 · persistence at 2 of 3 periods in the same direction · whole book at 1.0 point · negative variance beyond 1% under theoretical · and the Business Plan's whole-book policy of ±1.0 point or ±\$400, whichever is smaller, escalating at ±2.0 points sustained across two periods.
• Figure 34.6, three periods of meat and poultry, seafood, and produce, illustrating a pattern, pure volatility, and a chronic waste problem respectively.
• The meat-and-poultry decomposition: \$186 of stale cost cards · \$214 of uncosted pork special · \$92 of unchased credit · \$298 of ribeye over-portioning · \$207 of over-weight birds · \$997 explained · \$65 residual, which is 0.66% of ideal usage and inside a hand count's honest error.
• The two Run the Numbers calculations: the ribeye at 12.9 oz against a 12 oz spec, \$0.7763 a plate, \$298 a period and \$3,874 a year, closed by a scale already owned; and Chapter 15's over-pour at \$7,631 against \$233 of jiggers, a 32.7-times return.
• Figure 34.7, the investigation ladder, after Chapter 11's decomposition of a \$2,728 variance.
• Figure 34.8, dollars at risk against minutes of attention.
• Case Study 2 in its entirety — a constructed composite: a 96-seat restaurant with \$40,000 of period beverage sales, pour cost drifting 22.0% → 26.2% across seven periods, a Period 9 variance of \$1,680 decomposed to \$1,606 on rungs one through six and a \$74 residual, and \$20,920 of priceable cost incurred by starting at rung seven. No real business, employee, or event is described, and no figure in it should be cited as evidence about the industry.
• Case Study 1's arithmetic is Bellwether's; its historical and regulatory content is Tier 1 public record, and it explicitly declines to state any industry theft percentage.

Where to go next in this book

• Chapter 11 — Food Costing and Menu Engineering. Where the investigation ladder comes from. The \$2,728 decomposition that put theft seventh is the arithmetic this whole chapter rests on; if you read only one prerequisite, read that section.
• Chapter 13 — Purchasing and Inventory. Specs, par levels, the three-way match, and the highest-leverage twenty minutes of the day. §34.7 is the control layer over Chapter 13's procedure.
• Chapter 15 — Bar and Beverage Management. The five bar leaks and the \$16,169. This chapter owns the controls; Chapter 15 owns the leaks, the tenths-method count, and the pour arithmetic.
• Chapter 20 — Employment Law and Compliance. Punch edits, off-the-clock work, and the reason the highest-value permission decision in your point-of-sale system is a wage-and-hour decision.
• Chapter 21 — Leadership and Culture. Why the reliable people leave first, and the per-separation costs Case Study 2 uses to price a badly run investigation.
• Chapter 26 — Technology and the POS. The system of record, from which every exception report in §34.3 is drawn.
• Chapter 31 — Restaurant Accounting. The weekly flash report that seven daily sales reports add up to, and the comps, voids, and discounts lines this chapter reads as exceptions.
• Chapter 33 — Cash Flow and Working Capital. Where the difference between revenue and receipts becomes a survival question rather than a bookkeeping one.
• Chapter 35 — The Second Location. "Who audits the owners?" is a rhetorical question at one unit and a real one at two.