How to run a restaurant operations audit

An operations audit answers two questions. First: where is money leaking out of this business right now, and how much. Second: could someone who was not here at the beginning run this store correctly. Every finding should trace back to one of those.

That second question matters more than most operators expect. A restaurant can be profitable and still be unrepeatable, because the thing making it work is a founder who is there six days a week, or a kitchen manager who has been on the line for nine years and holds the entire system in her head. That is a fine way to run one store. It is why the third store fails.

An audit that only produces cost savings has done half the job. The other half is establishing what would have to be true for the business to exist without the people currently holding it together.

A complete audit covers seven areas. Skipping any of them tends to produce a report that recommends fixing a symptom.

• Financial performancePrime cost by store and by period, food and labor as a percentage of sales, average check, sales per labor hour, and the variance between locations. Variance is often the most useful number in the file: two stores running four points apart on the same menu is a systems problem, not a market problem.
• Menu and culinaryItem-level contribution margin against popularity, recipe adherence, plate cost versus theoretical, waste, prep efficiency, and whether the menu can actually be executed at volume by the staff you can realistically hire.
• Labor and schedulingDeployment against forecast, overtime, turnover by position, training time to competence, and whether managers are scheduling to demand or to habit.
• Supply chainDistributor agreements and rebates, price variance on the top thirty items, order guides, inventory discipline, and whether purchasing is centralized or happening store by store.
• Back of house flowStation layout, ticket times by daypart, throughput at peak, equipment fit, and the choke points that only appear on a Friday at seven.
• Guest experienceReview scores and their trend, complaint patterns, service standards as written versus observed, and speed of recovery when something goes wrong.
• Systems and documentationWhat exists in writing, whether anyone uses it, POS and reporting configuration, and how a new manager is currently brought to competence.

The sequence matters. Looking at numbers before seeing the operation produces confident conclusions about the wrong things.

1. 01

   Request the file

   Twelve months of P&Ls by location, current menu with recipes and costings, distributor agreements, org chart, any existing manuals, and review data. Read it before setting foot in a store.

2. 02

   Visit as a guest

   Unannounced, at a peak daypart, paying. This is the only time you will see the operation the way a customer does, and it is usually the most revealing hour of the whole audit.

3. 03

   Observe a full day

   Open to close, in the kitchen and on the floor. Prep, line, expo, service, close. Count what actually happens rather than what the manual says happens.

4. 04

   Interview by level

   Ownership, above-store leadership, general managers, and hourly staff separately. The gap between what leadership believes is happening and what the line cook describes is where most findings live.

5. 05

   Test the numbers against the floor

   Now reconcile. If theoretical food cost is four points below actual, the audit should be able to point at the specific stations and behaviors producing the difference.

6. 06

   Write it up ranked

   Findings ordered by financial impact and effort, not by category. The operator should be able to read the first page and know what to do on Monday.

The deliverable is not a presentation. It is a document the leadership team can work from for the next two quarters.

• A ranked findings listEvery finding with an estimated financial impact, the effort to fix, and an owner. Ranked by impact divided by effort so the sequence is obvious.
• A quantified opportunityWhat the operation should be producing at prime cost, and the gap to what it produces now, in dollars per store per year.
• A repeatability verdictA direct answer on whether the concept can be run by someone else, and precisely what is missing if it cannot.
• A ninety-day planWhat to fix first, who does it, and how you will know it worked. Without this the report becomes a shelf document.

Different brands, remarkably similar results. Across our engagements the same handful of issues account for most of the recoverable margin.

Purchasing left to the stores. Each location orders independently, nobody has renegotiated the distributor agreement in years, and rebates that were contractually earned are never claimed. This is frequently the single largest recoverable number and requires no operational change at all.

A menu nobody has costed since the last price rise. Ingredient costs move constantly; menus do not. It is routine to find items with negative contribution margin sitting in the most popular section of the menu.

Labor scheduled to habit. The same eight people on the same shifts regardless of forecast, because that is how it has always been. Deploying against actual demand typically recovers one to two points without reducing service.

Training that is really shadowing. New hires follow someone experienced for three days and are then declared trained. Nothing is tested, so standards drift a little with every hire, and nobody can point to when it started.

Off-premise treated as overflow. Third-party and catering orders handled on the same line as dine-in with no separate process, so margins are poor and dine-in service suffers whenever volume spikes.

There is no wrong time, but there are moments when the return is highest.

• Margins are drifting and you cannot say whySales are steady but profit is not. This is the classic case and usually the easiest to resolve.
• Before you open store number three or fourThe cost of finding out your systems do not travel is far lower now than after the lease is signed.
• Before considering franchisingThe audit is the first half of a franchise readiness assessment. Franchising without one is selling a system you have never verified.
• After a leadership changeA new operating executive inheriting a business needs an independent baseline, not a handover briefing.
• Before a raise or a saleInvestors will audit you. It is considerably better to have found the problems first.
• When performance varies wildly by storeVariance between locations running the same menu is almost always a systems failure and almost always fixable.