Key takeaways
- Raw sales and profit league tables reward the biggest site in the best location, not the best-run one.
- Normalise first: sales per seat, per opening hour or per square metre; labour as a percentage and as sales per labour hour; profit after a notional rent.
- Property costs vary by nation: Scotland, Wales and England each have different hospitality rates reliefs[1][2].
- Judge each site against its own history and a peer set of similar sites, and judge managers only on the lines they control.
Why raw league tables mislead
Once a business has two or three sites, the temptation is to rank them. The city-centre site takes the most money, so its manager must be the best; the market-town site makes the least, so it must be the problem. Often the reverse is true. A big site in a busy location can hide weak management behind footfall, while a small site with a thin catchment may be squeezing every pound out of its potential.
Unfair comparisons cause real damage. Good managers leave because they feel judged on things they cannot change, investment goes to the site that already looks strongest, and the group misses what the best-run site is doing right. The fix is to separate what each site is dealt (its size, location, lease and local market) from what its team does with it.
What differs between sites
- Size: covers, floor area, outside space and private dining rooms.
- Location: footfall, local incomes, office workers versus residents, tourism, parking and transport.
- Competition: how many other venues compete for the same guests nearby.
- Format: food-led or drink-led, all-day or evening only, takeaway and delivery share.
- Property: freehold or leasehold, rent level, rateable value and which nation's rates reliefs apply.
- Maturity: a site in its first year behaves differently from one that has traded for ten.
- Local labour market: how easy it is to recruit and what rates you need to pay.
Some of these can be measured and adjusted for. Others, such as local competition or labour markets, are better handled by grouping sites into peer sets and comparing within each set.
Normalise the numbers
| Raw measure | Fairer version | Why |
|---|---|---|
| Weekly sales | Sales per seat per week, or per opening hour | Removes the effect of size and trading hours |
| Covers | Seat utilisation: covers / (seats x services) | Shows how full the site is, not how big |
| Labour cost | Labour % of sales and sales per labour hour | Separates productivity from wage levels |
| Food cost | Actual vs theoretical food cost gap | Different menus have different target GPs |
| Site profit | Site profit after a notional market rent | Makes freehold and leasehold sites comparable |
| Reviews | Average rating and reviews per 1,000 covers | Busy sites collect more reviews regardless of quality |
| Growth | Like-for-like change against own last year | Removes the advantage of a newly opened site |
Use the gap between actual and theoretical food cost rather than food cost percentage alone. A steak-heavy menu at one site and a pizza menu at another will never have the same food percentage, but both kitchens can aim to be within a point of their own theoretical cost.
Property and rates: the nation matters
Groups that trade across the UK need to remember that business rates differ by nation. Rateable values from the 2026 revaluation are based on rental values at 1 April 2024[3]. In England, qualifying hospitality properties with rateable values under £51,000 pay the small business retail, hospitality and leisure multiplier, and those from £51,000 to £499,999 pay the standard one[4].
Scotland and Wales have their own schemes. In 2026-27, mainland Scottish licensed hospitality premises with rateable values of £100,000 or less receive an additional 25% rates relief[1], while Wales offers Food and Drink Hospitality Rates Relief of 15% of the bill, capped at £110,000 per business across Wales[2]. Two similar sites either side of a border can face very different rates bills, so compare them before rates, or after rates with a note explaining the difference.
Rates and rent are decisions made when you sign a lease, not by the site manager each week. Show them on the group's view of site profit, but leave them out of the manager's scorecard.
Competition density: same brand, different markets
Local markets vary enormously in how crowded they are. London had 20,955 restaurant and mobile food service enterprises in 2026, the most of any UK region, with the South East next on 13,275[5]. A site in central London competes with hundreds of venues within walking distance; a site in a Welsh market town may compete with a handful.
Counts like these are not a target for any single site, but they explain why a London site may need more marketing to hold its covers, and why a rural site's review score carries more weight with local guests who have fewer alternatives.
Pay is national, but labour markets are local
The legal minimum is the same across the UK: from April 2026 the National Living Wage for workers aged 21 and over is £12.71 an hour[6]. What you need to pay to recruit and keep good staff is not. A site in a tourist town in summer, or near a large employer that pays above the minimum, may need higher rates to fill a rota. When you compare labour cost between sites, look at sales per labour hour as well as the labour percentage, so a site paying higher wages but running efficient rotas is not penalised.
Maturity: new sites need their own yardstick
A new site goes through phases. Many open busy, as locals try them, then dip as novelty fades, then build steadily if the offer is right. Others start slowly and take a year to find their market. Neither pattern is a verdict on the manager in the first few months. Compare a new site against the opening plan you wrote for it: target covers, spend per head, labour and food cost by month. Bring it into like-for-like comparisons only once it has traded for a full year, and note any change of concept or major refurbishment as a fresh start.
Build a fair comparison in five steps
Group sites into peer sets
Put similar sites together: city centre, suburban, rural, drink-led, food-led. With only two or three sites, compare each mainly against its own history.
Choose normalised measures
Pick five or six from the table above and calculate them the same way for every site, with the same week definitions and cut-off times.
Compare trends before levels
Ask whether each site is improving against its own last year. For example, a small site growing 6% may be doing better work than a big one growing 1%.
Separate group and site decisions
Rent, rates, menu pricing set centrally and brand marketing belong to the group. Rotas, waste, service standards and local marketing belong to the site.
Review together
Share results in a monthly meeting where managers explain their numbers and swap what is working. A fair comparison is a learning tool, not just a ranking.
Example: three sites, one surprise
Example (fictional). A group runs a 140-cover city-centre restaurant, a 90-cover suburban one and a 60-cover village pub. Ranked on site profit, the city site comes first and the pub last. Normalised, the picture changes. The pub's seat utilisation is the highest of the three, its food cost is within half a point of theoretical against nearly two points over at the city site, and its like-for-like sales are up 5% against 1% in the city. Once a notional rent is added to the freehold pub, its profit per seat is close to the city site's.
The group's conclusion in this Example: the city site's manager has a cost control problem hidden by location, and the pub's head chef should be asked to share her ordering and prep routine with the other kitchens.
Common mistakes in group comparisons
- Ranking sites on total profit and handing out bonuses on that basis.
- Comparing food cost percentages across different menus instead of each kitchen's gap to its own theoretical cost.
- Ignoring different trading hours, so the site open seven days looks more productive than one open five.
- Forgetting that a central price change or marketing push affects some sites more than others.
- Changing the calculation from month to month, so trends reflect method rather than performance.
Manager scorecards: only what they control
- Labour against the agreed rota and sales per labour hour.
- Actual against theoretical food and drink cost.
- Waste recorded and waste trend.
- Review rating, complaints and how quickly they were answered.
- Food hygiene and health and safety checks completed on time.
- Staff turnover and training completed.
- Like-for-like sales against the site's own last year, with central price changes noted.
Frequently asked questions
Should site managers see each other's numbers?
Yes, if the numbers are normalised and you explain how they are calculated. Open comparison helps managers learn from each other. Raw league tables shared without context tend to demotivate the managers of smaller or harder sites.
How do I compare a new site with established ones?
Leave it out of like-for-like comparisons for its first 12 months and compare it against its own opening plan instead. After a year, compare it with its own previous year and with its peer set.
What is a notional rent?
An estimate of the market rent a freehold site would pay if it were leased, usually based on a surveyor's view or on comparable leases nearby. Adding it lets you compare freehold and leasehold sites on the same basis.
References
Scottish Government (2026). Non-domestic rates (opens in a new tab) ↑
Welsh Government (Business Wales) (2026). Non-domestic rates: Food and Drink Hospitality Rates Relief 2026-27 (opens in a new tab) ↑
HM Revenue & Customs and Valuation Office Agency (2026). Help with the 2026 business rates revaluation (opens in a new tab) ↑
HM Treasury (2026). Business Rates Multipliers: Qualifying Retail, Hospitality or Leisure (opens in a new tab) ↑
Office for National Statistics (2026). UK business: activity, size and location, 2026 (workbook) (opens in a new tab) ↑
GOV.UK (2026). National Minimum Wage and National Living Wage rates (opens in a new tab) ↑