The Room You Cannot Share
The data from this spring is not subtle. UK food halls grew from 114 to 149 venues in the twelve months to 2026. The development pipeline added another 25 percent. One venue closed. Meanwhile, conventional hospitality is losing somewhere between three and six businesses a day depending on which figures you use.
Two parts of the industry, moving in opposite directions, at the same moment.
What the food hall fixed
The food hall model is not complicated. Multiple independent food operators share one roof, one seating area, one central bar, one set of utilities, one front-of-house structure. The risk that would otherwise sit entirely on one operator's balance sheet is distributed across a building full of them.
For the chef testing a concept for the first time, the food hall is a way to trade without signing a lease, hiring a floor team, fitting out a kitchen and carrying all of it before selling a single cover. For the developer running the building, each stall generates footfall that makes the next stall more viable. The model makes economic sense because the capital and the cost are shared.
The surge in food hall openings is a rational response to the cost structure of hospitality. If fixed costs are the thing that closes restaurants, build something where the fixed costs are not fixed per operator.
What it cannot solve
The food hall route suits a specific kind of operation. Not a sitting-down, table-service restaurant with its own dining room, its own chef brigade, its own booking sheet and its own kitchen running a full menu. Not a neighbourhood pub that a community actually relies on. Not a hotel restaurant that exists to serve the rooms above it.
For the independent restaurant or pub that has made that commitment, there is no shared roof. The lease is signed. The kitchen is fitted. The rota is set on Sunday. The fixed costs are theirs entirely, and the room earns based on what comes through the door.
The food hall did not solve the problem of the quiet Tuesday. It offered a different model to the operator not yet tied to one. The operators already in a room are not in a position to restructure around shared infrastructure. They are in a position to fill the room they have.
The room that has already been paid for
We run the Castle Inn, Tollgate, the Bull, the Berwick, Ash and Honey, a tea room, a farm shop and a bakery. None of them have a shared seating hall or a distribution of fixed costs across five other operators. Each site carries its own lease, its own team, its own cost structure, whether the room fills or not.
That is not a complaint. It is the commitment that makes each of them what they are: a real place with its own character, its own kitchen, its own version of the same evening.
The fixed cost of that commitment runs whether the dining room is full or quiet. The kitchen prepped this morning for covers that may not all arrive. The rota was set before the booking sheet filled in. By 4pm, when the shape of the evening is clear, every cost in the building has already been committed.
The seat the food hall does not have
A food hall does not have an empty seat the way a restaurant does. It has footfall, or it does not. The cost of a quiet hour is distributed across the whole building, and someone at another stall is probably selling anyway.
A restaurant has a specific corner table, set and prepped and staffed around, that either earns something before service ends or earns nothing. The room cannot share that risk. It can only decide what to do with it.
Halfseat is built for exactly that decision. Around 4pm, venues release the tables they expect to lose: food at half price, drinks at full, a real cut of the booking fee going directly to the venue. Every pound at the bar stays there. The prep already happened. The rota was already set.
The seat that earns nothing is not a food hall problem. It is a real restaurant's oldest one.