Skip to content
Central Production Kitchen: Build, Buy, or Outsource?
There are three ways to get centrally produced food: build your own production kitchen, buy finished dishes from someone else's, or hand the whole thing to a contract caterer who brings their own. Building wins when you have enough sites and steady enough volume to keep the plant busy in your quietest month. Buying converts capital into a per-portion cost with no compliance burden. A contract caterer removes the operational load entirely, in exchange for a management fee — typically 8–15% — and a multi-year commitment.
Why the model keeps coming up
Centralising production is the standard answer to the standard problem: cooking from raw at every site, to a consistent standard, needs skilled people at every site. Cook once in one place, distribute, and regenerate at the point of service, and one team's skill reaches every location.
That logic is sound, and it is why the model has spread from hospitals and school caterers into hotel groups, workplace operators and multi-site leisure. The question is not whether centralised production works. It is who should own the plant, the people and the risk — and there are three answers, not two.
Option 1 — Build your own
The capital is the visible part: building or fit-out, blast chillers or freezers, production equipment, cold storage, distribution vehicles. It is rarely what catches people out.
The harder parts are these. Utilisation: a production kitchen is only economic near capacity, and demand across your sites has to be steady enough to keep it there. Compliance: you now operate a food manufacturing site, with the HACCP regime, traceability, temperature records, allergen control and audit exposure that implies — materially different from running a service kitchen. Recruitment: you still hire and keep a production team, only now the entire estate fails if that one team fails. Distribution: vehicles, routes, temperature control, and their fixed cost whether they run full or not.
You have concentrated labour risk rather than removed it. For a large estate that is a good trade — one strong team is easier to run than fifteen weak ones. For a small one it is a single point of failure with a mortgage.
Option 2 — Buy finished dishes from a production kitchen
Buying gets you the output without the plant. Dishes are developed and cooked by chefs, blast-frozen at peak quality, and regenerated on site in a smart oven — about seven minutes for an individual portion, trays serving five. Around twelve months of frozen shelf life means delivery is roughly monthly rather than a daily distribution run of your own.
Capital is zero, compliance sits with the supplier, and cost is variable: you pay per portion, so quiet months cost less instead of exposing an underused asset. At 100 covers, £3.10–£4.50 per meal on two staff, with no management fee.
What you give up is authorship. You work from a menu library, and you depend on a supplier. If your estate is large enough to run its own plant well, that dependency may not be worth accepting.
Option 3 — Let a contract caterer run it
The third option is the one most often left out of a build-versus-buy paper, and it is the one large operators most often actually choose. A contract caterer — Sodexo, Compass, Aramark, Elior, BaxterStorey and others — brings its own production, its own staff and its own supply chain, and takes the catering operation off your desk entirely.
The appeal is genuine. You transfer operational responsibility, employment risk and food-safety accountability in one move, and you get a single accountable partner across every site. For an operator whose core business is not food, that is worth paying for.
The cost is a management fee, typically 8–15%, and usually a multi-year commitment with meaningful exit terms. You also lose visibility: the margin, the specification and the day-to-day decisions now sit with a third party, and quality tends to vary site by site with whoever they have staffed you with. It is the most hands-off option and the least controllable one.
Build, buy, or outsource?
| Build your own | Buy finished dishes | Contract caterer | |
|---|---|---|---|
| Capital | Significant, up front | None | None |
| Cost behaviour | Largely fixed — quiet months still cost | Variable, per portion | Fee-based, typically 8–15% management fee |
| Compliance burden | Yours — a food manufacturing site | Supplier's | Caterer's |
| Labour risk | Concentrated in one team the estate depends on | Removed from your side | Transferred to the caterer |
| Control | Complete | Menu library, but you keep the operation | Lowest — spec and margin sit with a third party |
| Commitment | The asset's life | Order to order | Multi-year, with exit terms |
| Best when | Many sites, steady high volume, long horizon | Fewer sites, variable demand, or you want to start now | Food is not your core business and you want it off your desk |
The threshold questions
Before modelling anything, answer this: how many portions a week would the kitchen produce in your quietest month, not your busiest? Production kitchens are sized by peak and killed by trough. If the quiet-month number will not keep the plant reasonably busy, the business case is weaker than the spreadsheet suggests.
Second: what happens if the production team walks? In a decentralised estate that is one site having a bad week. In a centralised one it is every site, on the same day.
Third, and the one that separates buying from outsourcing: do you want to keep the catering operation, or hand it over? Buying keeps you running the service with a supplier behind you. A contract caterer takes the service itself. Those are different businesses to be in, and the fee is the price of not being in one of them.
A staged approach is often sensible — buy in while volume builds, and revisit building once the quiet-month figure would keep your own plant busy. It also gives you a real operating benchmark to model a build against instead of a projection.
- There are three routes to centrally produced food, not two: build your own plant, buy finished dishes, or hand the operation to a contract caterer.
- Building means running a food manufacturing site — HACCP, traceability, allergen control, audit exposure — and it concentrates labour risk into one team the whole estate depends on.
- Buying converts capital into a variable per-portion cost and moves compliance to the supplier, while you keep control of the service.
- A contract caterer removes the operation entirely for a management fee of typically 8–15% and a multi-year commitment, at the cost of control and visibility.
- Size any build on your quietest month — production kitchens are killed by low utilisation, not by low headline volume.
Frequently asked questions
A single facility that cooks food for multiple sites. Dishes are produced in bulk, chilled or blast-frozen, distributed, and regenerated at the point of service, so culinary skill is concentrated in one place instead of required everywhere.
When enough sites and steady enough volume would keep the plant near capacity in your quietest month, and the time horizon justifies the capital. Utilisation, not headline volume, decides it.
A contract caterer takes over the catering operation itself — their staff, their supply chain — for a management fee of typically 8–15% and a multi-year term. Buying finished dishes leaves you running the service, paying per portion, with no fee and no lock-in.
It is food manufacturing rather than food service: a full HACCP regime, traceability, temperature and process records, allergen control and audit exposure across everything you distribute.
It concentrates it. You need fewer skilled people overall, but the whole estate now depends on one production team — better for a large estate, riskier for a small one.
Yes, and it is often the sensible sequence. Buying in while volume builds gives you real operating data to model a future build against, rather than a projection.
Related reading
Contact Us : novachef.co | info@novachef.co | +44 20 805 87184