Centralized Production Kitchens: Financial and Operational Viability Analysis of the Commissary Model
Straight verdict: a commissary stops being optional once a group crosses 8-10 units and its food cost variance across locations climbs several percentage points. Below that threshold, the CapEx of a central plant destroys margin instead of creating it. This is not a culinary decision: it is a unit economics decision. Centralizing turns food cost into a purchasing-and-process variable, controllable and negotiable at volume, instead of 10 chefs improvising the same recipe with 10 different waste curves. The mistake I see over and over is building a commissary to show off installed capacity rather than to cut prime cost. If the plant doesn't push theoretical versus actual cost below 32% food cost per dish and doesn't recover its CapEx in 24-36 months, it isn't infrastructure: it's a liability with LED lights.
Ten independent kitchens turn out ten versions of the same dish, ten waste curves, ten actual food costs that rarely match theory. Every new unit replicates the concept and the variance along with it. The commissary (a centralized production kitchen) promises to close that gap by turning production into a controllable industrial process. The price: heavy CapEx and a different operating architecture.
One question drives this white paper, answered from the cash register: at what point does a commissary stop being a capital burden and become a margin lever? It is not a chefs' debate. It is unit economics, operational maturity and food cost variance risk at scale. Segment (QSR, fast casual, full service), unit count and input inflation all move the answer, so each chapter breaks down along those three dimensions.
Side-by-side: centralized production kitchen
| On-Site Kitchen (decentralized) | Commissary (centralized production) | |
|---|---|---|
| Initial CapEx per production unit | ✕Bundled into the unit fit-out cost, depending on the QSR format and location. | ✓One-time concentrated CapEx: varies significantly by capacity and equipment. |
| Food cost variance across locations | ✕High: 3-6 pt typical dispersion between units | ✓Low: a fraction of a point; the central process standardizes theoretical vs. actual cost. |
| Prime cost (food + labor combined) | ✕Full service is well above that range; it rises with every kitchen doubling stations. | ✓Dilutes production labor; concentrates volume and lifts buying power |
| Sustainable net margin by format | ✕Full service runs on the sector's tightest net margin; fast casual holds the higher ceiling, per Peppr POS 2025. | ✓Enables the top of the range by cutting waste and standardizing purchasing |
| Supplier negotiating power | ✕Fragmented: each unit buys its volume separately | ✓Consolidated: volume buying, short supply chains |
| Model profitability threshold | ✕Optimal up to 6-8 units or low recipe complexity | ✓Profitable from 8-10 units with process-intensive recipes |
| CapEx recovery (payback) | ✕N/A (distributed cost) | ✓24-36 months if food cost drops 2-4 pt on consolidated sales |
Chapter 1 — At what point does a commissary make sense?
A commissary stops being optional once a group crosses 8-10 units and food cost variance across locations climbs several percentage points. Below that, the CapEx doesn't clear:
a central plant starts where a large unit's fit-out ends. I've seen it in dozens of groups. Six kitchens can still be tamed with audits and a standard recipe; the eighth brings its own waste curve, the ninth another. At that scale, one loose point of food cost, multiplied across ten registers, eats a fast casual's already thin margin.
Chapter 2 — A commissary turns food cost into an input variable
Centralizing has one concrete financial reason: food cost stops being a result and becomes an input variable. You audit one process, not ten kitchens. The mistake I see again and again is measuring waste at closing; a commissary buys and processes to a fixed spec, so output is standardized from birth. Little room for error here, with a well-run fast casual's net margin capped at 9% per Peppr POS 2025.
Ten units buying separately never negotiate like one plant consolidating volume, and that is where short supply chains are born. Fail to industrialize production at that scale and you hand competitors margin they already capture through purchasing power.
Chapter 3 — Concentrated CapEx is an accounting objection, not an operational one
Boards rarely object to a commissary's price tag; they object to the accounting exposure of an asset that takes 24-36 months to pay off. On-site, CapEx hides in each unit's fit-out. Centralized, it lands in one outlay someone has to sign. For scale, a full McDonald's franchise runs well past a million dollars in initial investment, and a Burger King sits in a similar range.
A serious commissary sits between one and two premium units of that kind. As a Masterestaurant consultant, Diego F. Parra frames it for the board this way: not an expense, but ten scattered operational risks converted into one controllable asset, with returns measured in closed food cost variance.
Chapter 4 — Purchasing power scales differently than operations
Ten units buying separately never negotiate like one plant consolidating their volume: purchasing power scales non-linearly. That marginal efficiency, the cradle of short supply chains, is what the decentralized model never reaches. On that cost base, one point of food cost won through centralized buying beats any promotion. Chains chasing scale know it. Chipotle targets 7,000 restaurants in North America per Restaurant Dive 2025; Wingstop, 10,000 units worldwide.
Nobody gets there with ten kitchens buying alone: they get there because a central plant fixes spec and price, and nails down waste before the input ever reaches the unit.
Chapter 5 — How does the threshold change by segment and inflation?
The threshold falls with input inflation and rises with menu fragmentation. In QSR, production is repetitive by design, and a commissary pays off fast even at six units.
Full service is another story: at a noticeably lower net margin per Peppr POS 2025, a broad menu dilutes the savings and break-even drifts toward more units. Fast casual, high volume on a tight menu at a margin ceiling near 9% per Peppr POS 2025, is the ideal ground. And if inflation spikes? Every point of variance costs double, the plant's savings grow in step, and break-even moves forward six to nine months. One rule fits nobody, which is why this analysis splits the threshold by segment and price scenario.
Chapter 6 — The real case: closing the gap between theoretical and actual food cost
Closing the gap between theoretical and actual food cost is the commissary's job: in decentralized groups that gap opens several percentage points between best and worst unit. I've audited groups where theory and actual food cost diverged by several points on a fast casual margin that, per Peppr POS 2025, tops out near 9% — enough to separate growth from closure.
And the market has gone K-shaped: the largest chains kept growing while the next tier fell back, per Technomic 2025. The operating difference between them is process control, nothing more. A central plant doesn't guess waste; it fixes it on the line, spec and weight controlled before dispatch to each unit.
Chapter 7 — When does a commissary destroy margin instead of creating it?
A commissary destroys margin when built too early or oversized: a plant running well under capacity loads fixed cost onto too few registers and sinks the unit economics.
Here sits the model's paradox: the same infrastructure that promises margin to ten units eats it at five, because savings per unit grow with volume and fixed cost waits for no one. The fix is a hard register-side rule: don't centralize until you can fill most of the line, and compute break-even without loading payroll or rent onto the plate; those belong to the group's break-even, not per-dish food cost.
Oversizing is expensive when a franchise's initial investment already starts in six figures. Starbucks runs 38,587 locations per Restaurant Business 2024 on a logistics network that only closes with density. The Masterestaurant lesson: a margin lever once scale pushes variance out of control; before that, an idle asset the balance sheet won't forgive.
Chapter 8 — The differences the board decides, not the chef
A commissary changes what food cost is: an outcome (what came out) becomes an input variable (what you buy and process). That is the real financial case for centralizing. Auditing 10 kitchens turns into controlling 1 process. On-site kitchens spread CapEx across units and hide it in the fit-out; a commissary concentrates it in one visible outlay the board has to approve. The objection is rarely the amount.
It is the accounting exposure of an asset that takes 24-36 months to pay off. Buying power scales differently. Ten units purchasing separately never negotiate like one plant consolidating volume; that consolidation is where short supply chains are born, along with a marginal efficiency the decentralized model can't reach.
On-site vs. commissary: criterion-by-criterion analysis
When the on-site kitchen still wins
- Groups of 1 to 6-8 units: commissary CapEx isn't amortized by the volume generated yet.
- Simple-execution concepts (assembly, few production SKUs): centralized waste control adds little.
- Markets with fragile cold logistics or long distances: distribution cost eats the purchasing savings.
- Premium full service where 'made fresh on-site' perception is part of the average ticket.
When the commissary becomes mandatory
- 8-10+ units with food cost variance clearly diverging across locations.
- High-turnover QSR and fast casual: the centralized process stabilizes theoretical vs. actual cost.
- Process-intensive recipes (stocks, mother sauces, baking, marinades) that multiply waste in open kitchens.
- Aggressive franchise scaling plans where the replicable operations manual demands identical product.
Figures that frame the decision (sources 2024-2026)
“A fast casual group with 11 units had 31.4% food cost in its flagship and 37.8% in the most distant one: 6.4 points of dispersion no one saw in the consolidated number. We centralized mother sauces, marinades and portioning in a USD 780,000 commissary. Within 14 months the variance fell to 1.2 points and consolidated food cost dropped from 34.9% to 31.6%.
On annual sales of USD 9.2M, those 3.3 points are USD 303,000 a year. The CapEx paid off in 31 months. It wasn't culinary magic: it was turning 11 waste curves into 1 measurable process.”
Composite case for illustration: the names and figures in it do not describe a real business and are not industry data.
90-day roadmap to decide and build the commissary
Measure the actual food cost of EACH unit for 4 weeks and calculate the dispersion (food cost variance) against the theoretical cost. If variance across units keeps diverging, the commissary case exists. In parallel, run territorial prefeasibility with location intelligence: site the plant minimizing cold-chain miles to the location network. Without this diagnosis, you build installed capacity blind.
Build the commissary's unit economics: equipment and cold-chain CapEx, incremental OpEx (production payroll, logistics) and expected savings in food cost and volume buying. For example, simulate a handful of input-inflation scenarios and verify payback lands within a reasonable window in the base case. If it doesn't cut food cost at least 2 points on consolidated sales, the model doesn't close.
Translate every mother recipe into a process sheet with yield, target waste and critical control point. This replicable operations manual is what makes 1 plant produce what 10 kitchens produced with 10 criteria. Define theoretical cost per SKU and the daily actual-cost measurement system. It is the foundation of food cost variance control at scale.
Start by supplying 2-3 representative units (not the whole network). Measure for 3 weeks: food cost variance, prime cost, actual vs. target waste, logistics on-time delivery and unit operational satisfaction. Adjust routes and portioning before scaling. The pilot catches most of the cold-logistics problems a full rollout would make you pay for dearly.
And with AI?
Standardize and replicate processes to scale and franchise with control. Diego F. Parra is an expert in AI applied to restaurants.
Centralized production kitchen: free tools to start today
Masterestaurant ecosystem tools for this decision
Deciding on a commissary is a unit economics exercise, not intuition. These three tools from the Masterestaurant method turn the debate into numbers the board can approve.
Use them in sequence: first the business model, then scaling and finally the cash flow that sustains the CapEx through the payback period.
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.
Frequently asked questions from a CFO or Expansion Director
What is the kitchen expansion cost of a commissary versus on-site kitchens?
What is the kitchen expansion cost of a commissary versus on-site kitchens?
On-site, kitchen expansion cost is spread across units and hidden in each location's fit-out; a commissary concentrates it in one outlay that varies significantly by capacity and equipment. It only pays off if it cuts food cost and recovers that investment within a reasonable payback. With few units, building another full kitchen is usually cheaper than a central plant; once food cost variance across locations climbs, the commissary turns scattered operational risks into one controllable asset.
What is a commissary kitchen?
What is a commissary kitchen?
A commissary kitchen is a central production plant that buys, processes and standardizes prep for several units of a restaurant group, so every location receives product to the same spec. Food cost stops being each kitchen's result and becomes a purchasing-and-process variable audited in one place, with volume buying power the individual units never reach.
When does a central kitchen beat on-site kitchens?
When does a central kitchen beat on-site kitchens?
A central kitchen beats on-site kitchens once a group runs enough units that food cost variance across locations climbs, especially with process-intensive recipes. Below that scale, audits and a standard recipe still tame variance, and the CapEx of a central plant destroys margin instead of creating it.
How much does a commissary kitchen cost?
How much does a commissary kitchen cost?
How much a commissary kitchen costs varies widely with capacity and equipment, and it lands as one concentrated CapEx outlay; a serious plant sits between one and two premium franchise units' worth of initial investment. It only makes sense if it pays back within a few years by cutting food cost on consolidated sales.
What is a central production kitchen and when does it make sense for a restaurant group?
What is a central production kitchen and when does it make sense for a restaurant group?
A central production kitchen, or commissary, is one plant that buys, processes and standardizes prep for several locations to a fixed spec, so food cost stops being each kitchen's result and becomes a controllable input variable. It makes sense once a group has enough units that food cost variance across locations costs more margin than the plant's CapEx. Below that threshold, audits and a standard recipe in each unit remain the more profitable choice.
What is a commissary model for restaurants and how does it work?
What is a commissary model for restaurants and how does it work?
A commissary model is a central kitchen that buys, preps and portions food for several locations, which then only finish and serve it. It runs on fixed specs: every recipe leaves the plant standardized, so food cost stops depending on each chef and becomes a purchasing-and-process variable you audit in one place. It consolidates volume with suppliers and narrows waste across units, but concentrates CapEx in one large outlay, so it pays off once food cost variance between locations can no longer be fixed with audits and standard recipes.
From how many units is a commissary worth it?
From how many units is a commissary worth it?
As a rule, from 8-10 units with food cost variance running notably high across locations. Below that, the CapEx isn't amortized by the volume generated. Recipe complexity can lower that threshold.
How long to recover the investment?
How long to recover the investment?
In a well-sized model, CapEx payback lands at 24-36 months, provided centralization cuts food cost 2-4 points on consolidated sales and consolidates buying power. If it doesn't reduce food cost by at least 2 points, the model doesn't close and the asset becomes a liability.
Does a commissary improve net margin or just stabilize it?
Does a commissary improve net margin or just stabilize it?
Both, but it stabilizes first. It closes food cost variance dispersion and then, via volume buying and lower waste, enables the top of the margin range by format: up to 9% in fast casual versus a lower range in full service, per Peppr POS 2025.
What's the biggest risk of centralizing production?
What's the biggest risk of centralizing production?
Cold logistics. A poorly sited commissary or one with long distances turns purchasing savings into distribution cost and cold-chain break risk. That's why territorial prefeasibility with location intelligence is step 1, not a later detail.
2026 data on centralized production kitchen
Verifiable industry benchmarks from official, non-commercial sources (government, industry associations, market research) - not competitors.
| Metric | Value | Source |
|---|---|---|
| cumulative closure at three years: the real filter sits in year two, not year one | 60 percent (total failure rate for the three-year period) (2005) | Cornell University (Cornell Hotel and Restaurant Administration Quarterly, hoy Cornell Hospitality Quarterly) — Cornell Quarterly Article Examines Restaurant Success Rate 2005 |
| of independent restaurants close during year one — not the 90% of the myth | 26% (26.16% first-year failure, per the original study by Parsa, Self, Njite and King) (2005) | Cornell Hotel and Restaurant Administration Quarterly (H.G. Parsa et al.) — Cornell Quarterly Article Examines Restaurant Success Rate 2005 |
| EU urban population exposed to PM2.5 concentrations above the WHO annual guideline level (5µg/m³), 2022 data | 96% (of the EU's urban population exposed to PM2.5 concentrations above the WHO annual guideline level of 5µg/m³, 2022 d | European Environment Agency — Europe's air quality status 2024 |
| Payroll (salaries and benefits) as a share of sales in full-service restaurants | 36.5% of sales in 2024 (median, full service); ~33% historical average in the 2010, 2013 and 201 | National Restaurant Association — Restaurant labor costs are well above historical averages (Restaurant Economic Insights, Analysis & Commentary) 2025 |
| Of sales goes to food and beverage cost, segmented by restaurant type and sales volume (not a single 'average restaurant' figure) | 32.4% (limited-service, median) / 31.0%-33.7% (full-service, by sales volume) | National Restaurant Association — Higher volume restaurants reported lower food-cost ratios in 2024 (Restaurant Operations Report / Restaurant Operations Data Abstract) |
| Columbus, Ohio restaurants (1996-1999) that closed during their first year of operation, per H.G. Parsa's longitudinal study | 26 percent (first-year failure) (2003) | Ohio State University (news.osu.edu), on H.G. Parsa's longitudinal study: Restaurant Failure Rate Much Lower Than Commonly Assumed, Study Finds 2003 |
Download this document as PDF
The full text is free to read on this page. To take the corporate PDF with you, leave your details — we'll also email you the direct link.
Related content
The Masterestaurant method for centralized production kitchen
Applied in +8.400 restaurants across 43 countries.