Industry POV
Four-wall analysis: what a site earns before head office touches it
Four-wall profit is what a location makes counting only the revenue and costs inside its own four walls. How to calculate it, what it is for, and the mistake that makes it flattering.
Harry Soar23 August 20264 min read
Four-wall analysis measures what a single site earns using only the revenue and the costs that happen inside that site. Head office comes out. Central marketing comes out. Interest, depreciation and the finance director come out. What is left is the number that answers one question: does this location make money on its own terms.
It is the standard way multi-site operators compare venues that are not otherwise comparable, and it is the number a buyer or a lender will ask for first. It is also frequently explained wrongly, including in the version of this article we used to run, so it is worth being precise.
What goes in, and what stays out
In: everything the site generates and everything it consumes.
- Site revenue, all channels. Counter, table, your own online ordering, the marketplaces, click and collect.
- Cost of goods for what that site sold.
- Site labour, including the general manager and their national insurance and pension.
- Rent, rates, service charge, utilities.
- Site-level repairs, cleaning, consumables, waste, card fees and marketplace commission on that site's orders.
Out: anything that would still exist if you closed the site tomorrow.
- Head office salaries, the support centre, the finance and marketing teams.
- Group marketing, brand spend, the website.
- Interest, depreciation, amortisation, corporation tax.
- Central software licences, unless you can honestly attribute them per site.
That last exclusion is the point of the whole exercise. Overhead allocated across sites tells you about the allocation formula, not about the site. A venue can look poor because it carries a large share of central cost and good the moment you take that share away, and the operator needs to know which of those two pictures is real before deciding anything.
How to calculate it
Take one site, one period, and work down:
- Net sales for the site, across every channel, excluding VAT.
- Less cost of goods sold, to get gross profit.
- Less site labour, to get what most operators call contribution.
- Less occupancy and site operating costs, to get four-wall profit.
Express it as a percentage of net sales as well as a cash figure. The percentage is what makes two sites of different sizes comparable, and the cash figure is what tells you whether the percentage is worth anything.
What it is actually for
Three decisions, mainly.
Which sites to fix. A four-wall margin well below the estate average is a site problem, and it is usually labour scheduling or waste rather than sales. It gives you somewhere specific to look.
Whether to keep a site. A location that is four-wall negative is consuming cash to stay open, and closing it improves the group immediately. A location that is four-wall positive but loses money after overhead is a different problem entirely, and closing it makes the group worse, because the overhead does not leave with it.
Whether the next site is worth opening. Four-wall margin at maturity, against build cost, is the return on a new opening. It is the number that decides whether growth is growth or just more sites.
The mistake that makes it flattering
Four-wall profit is not company profit, and it is not free cash. Add every site's four-wall profit together and you have not arrived at what the business made, because head office is still unpaid. An estate where every site is four-wall positive and the group still loses money is common and is not a contradiction: it means the overhead is too large for the estate, which is a head office decision rather than an operations one.
The other one is inconsistency. If one site's manager sits in site labour and another's is paid centrally, the two four-wall numbers are not comparable and the comparison is worse than useless because it looks rigorous. Write the rule down once, apply it to every site, and keep it stable across periods.
Getting the numbers out in the first place
Most of the difficulty is not the arithmetic. It is that site revenue now arrives through several systems that do not agree with each other, and the commission, refunds and channel fees sit somewhere else again. If your marketplace sales land as a single monthly settlement figure, the site's real gross profit is not visible.
What makes the analysis routine rather than an annual project is having sales attributed to the site at the moment they happen, split by channel, net of what each channel charges. That is a reporting requirement more than a software feature, and it is worth specifying it before you buy anything. Our multi-site setup is built around it, order and pay keeps in-venue sales attributed to the right site and server rather than to a shared till, and the integrations list covers what pushes into the accounting and EPOS systems that hold the cost side.
Questions operators ask
What is four-wall analysis in a restaurant?
It is a measure of what a single site earns using only the revenue and costs generated inside that site. Head office, group marketing, interest, depreciation and tax are excluded, so what remains reflects the location's own performance rather than how central overhead was allocated to it.
What is included in four-wall costs?
Cost of goods sold, site labour including the general manager, rent, rates and utilities, and site-level operating costs such as cleaning, waste, repairs, card fees and marketplace commission on that site's orders. The test is whether the cost would disappear if the site closed.
Is four-wall profit the same as EBITDA?
No. Four-wall profit sits above it. EBITDA for the group takes the combined four-wall profit of every site and then subtracts head office and central costs, so a business can have healthy four-wall margins at every location and still make a loss overall.
Should you close a site with a low four-wall margin?
Only if it is four-wall negative, meaning it consumes cash to stay open. A site that is four-wall positive but unprofitable after overhead is still contributing towards costs that do not go away when it closes, so closing it usually makes the group worse rather than better.
More from the blog
Choosing delivery apps
Work out whether a marketplace order makes you money before you argue about which one to join. The arithmetic, the marketing argument that usually is not one, and where a direct channel fits.
Harry Soar · 23 August 2026 · 4 min read
Read post →
Just Eat alternatives
Four different things get called a Just Eat alternative and they solve different problems. Which one you want depends on whether you are replacing the demand, the delivery, or the ordering.
Harry Soar · 23 August 2026 · 3 min read
Read post →
EPOS for takeaways
The three systems most takeaways look at first came out of card readers, not kitchens. What that means in a rush, and what to compare before you commit.
Harry Soar · 23 August 2026 · 2 min read
Read post →
Ready to
move?
Free to set up, live the same day, and your guest list comes with you.