New product development

What a sprint delivers

1. A written development brief before your kitchen cooks anything: target plate cost, target margin, prep time ceiling, skill level, equipment constraints and supply requirements.
2. A dish count agreed in the brief, each one costed, specified and photographed.
3. Full specifications written for a kitchen you have never met: quantities in grams, method in steps, hold times, plate up and yield.
4. Supply validation. Every ingredient checked for availability at your volume, through your existing suppliers, at a price that holds.
5. A rollout trial in a representative sample of your sites before anything goes estate wide.
6. Training material your development chefs can actually deploy: one page per dish, photographs, common failure points.

A dish that has to travel is a different dish

Developing one dish for one kitchen is cooking. Developing a dish that has to come out the same in eight kitchens, with eight sets of equipment and eight different standards of section chef, is a specification problem that happens to involve food.

The failure mode is familiar. A dish is developed by a talented development chef in a well equipped test kitchen with time and attention. It is beautiful. It goes to the estate. Within six weeks it has quietly become four different dishes, none of which is the one that was costed, and the plate cost nobody is checking has drifted by well over a point.

So the constraints go into the brief before anyone turns a pan on. Target plate cost. Target margin. A ceiling on prep time, which is a cost in its own right and the subject of this guide on menu labour. The skill level of the section that will actually cook it on a wet Tuesday, not the one in the test kitchen. The equipment every site genuinely has, which is rarely the equipment the estate thinks it has. Developing inside those constraints is harder and slower, and it is the only way the dish that launches is the dish that was signed off.

Specification is the whole job

A specification written for a kitchen you have never met cannot assume anything. Quantities in grams, not handfuls. Method in numbered steps. Hold times and what the dish looks like when it has been held too long. Yield, so the plate cost survives contact with portioning. Plate up, with a photograph taken from the angle the chef will be looking at it.

It also has to name the failure points, which is the part most specs skip. Where does this dish go wrong? Which component is the one that gets rushed when the section is buried? What does it look like when the sauce has split, and what does the chef do about it at eight o'clock with no time to start again? A development chef knows these things. If they are not written down, they stay in one head and never reach the site that needs them.

Supply is where estate dishes die

Every ingredient gets checked against the volume you will actually order, through the suppliers you actually use, at a price that holds for the life of the menu.

This sounds procedural and it kills more estate dishes than anything in the kitchen. A garnish available in boutique quantity through a specialist is not available for forty sites. An ingredient on a seasonal price that looks workable in October is a different proposition in February. A single source item with no second supplier is an outage waiting to happen, and when it happens the sites substitute something, nobody records what, and the costing is fiction from that day on.

Anything that cannot be supplied at volume, through your existing route, at a price that holds, does not go in the brief. It is cheaper to find that out during development than during rollout.

The rollout trial

Before anything goes estate wide it runs in a representative sample of sites. Representative means the awkward ones as well as the flagship: the smallest kitchen, the one with the oldest equipment, the one with the highest turnover of section chefs.

A dish that holds in those sites will hold anywhere in the estate. A dish tested only in the sites that cook everything well tells you nothing you needed to know.

What comes back from the trial is the training material: one page per dish, photographs, the failure points, and whatever the trial kitchens discovered that the specification had not anticipated. They always find something.

What the fee depends on

Quoted per sprint rather than from a price list. The number is driven by the number of dishes in the brief and the number of sites the specification has to hold across, since a dish that works on one section is a different problem from one that has to travel.

Timing is set per sprint against your own launch date. Two things drive it: the dish count, and how long your kitchens need to trial properly, which is the step that cannot be compressed. For a single new site the same sequence, at a smaller scale, is set out in how to open a restaurant from the kitchen side. Most operators run sprints against their seasonal change cycle and brief well ahead of the intended launch. You get the figure, and the brief, before development starts.

Where sprints go wrong

Developing to a flavour brief with no commercial constraints, and discovering the plate cost afterwards. By then the dish has a champion and taking it off the list is a political problem instead of an arithmetic one.

Testing only in the good kitchens. Rolling out without training material, so each site interprets the dish. And briefing too late, which compresses the trial, which is the one part of the process that cannot be compressed without turning the rollout into the trial.

This is also not a creative consultancy engagement. If what you want is someone to invent exciting food with a free hand, the constraints here will frustrate you. Every dish is developed against a number, and the number is agreed first. If you have an existing menu that needs fixing instead of extending, an audit is the cheaper starting point.

A new dish is judged by the section it joins. See what it has to earn in main course profitability and dessert menu profitability, and how a starter changes the shape of the order in starter menu optimisation.

New product development FAQs