Most restaurant groups track labour as a percentage of revenue. The number appears every week on the P&L, sits alongside food cost, and gets reviewed in the Monday call.
It feels useful but…
The problem is that it is measuring the wrong thing – and in multi-site groups, the distortion compounds in ways that make operational decisions harder, not easier.
Here is the issue, and what to do about it.
The standard calculation and why it fails
The typical labour percentage calculation is straightforward: total labour cost divided by net revenue, expressed as a percentage.
For example: if you spent £40,000 on labour last week and took £200,000 in revenue, your labour percentage is 20%.
The problem starts the moment you run two sites with different revenue profiles:
- Site A takes £150,000 a week with a lean team.
- Site B takes £80,000 with a similar headcount because it is smaller and newer.
Site B’s labour percentage will almost always look worse, even if the team there is more efficient, better scheduled, and tighter on hours.
You are comparing a percentage against an absolute cost base that is partly fixed.
Rent, management, kitchen structure – these do not scale with revenue the way variable labour does.
When revenue is lower, fixed labour drags the percentage up.
When revenue is higher, it hides inefficiency because the denominator is doing the work.
The percentage tells you what happened.
It does not tell you why, or where.
What actually drives your labour cost
Labour cost in a multi-site hospitality group is driven by four things:
- Scheduled hour
- Actual hours worke
- Wage rate by rol
- Revenue per labour hou
Most groups track the first two reasonably well through their rota software.
Fewer track wage rate by role with enough granularity – the blended average hides the mix shift between senior and junior team members.
Almost none track revenue per labour hour consistently, which is the number that actually tells you whether a site is efficient.
Revenue per labour hour is simple: total revenue divided by total hours worked.
If site A generates £45 per labour hour and site B generates £28, that is a meaningful operational difference that your labour percentage may not surface – especially if site B has higher revenue and therefore a lower percentage.
Where multi-site groups lose money
The most common loss points we see in groups running 4 to 15 sites are:
- Overtime that accumulates invisibly.
Staff who are regularly scheduled at 39 hours but clock 44.
Across 10 sites, this can be worth £1,500 to £3,000 per week and never appears as a line item.
It is buried in the blended labour number.
- Management payroll misallocated across sites.
A general manager who covers two sites, or an area manager whose cost sits centrally rather than by site, makes site-level labour reporting meaningless.
You cannot manage a number you cannot attribute accurately.
- Seasonal ramp-up not modelled against forecast.
Most groups increase headcount when they get busy.
Fewer cut it with equal precision when revenue drops.
The labour percentage only reflects this after the fact, when the period is closed and the money is spent.
- Salaried versus hourly not separated.
Mixing salaried management costs with hourly team costs in the same labour line produces a number that responds to revenue changes in ways that do not reflect operational reality.
A 2% improvement might be entirely driven by a good revenue week rather than any change in team efficiency.
What to measure instead
The metrics that actually give you operational leverage at group level are:
- Revenue per labour hour, by site, by day part.
This tells you which sites are efficient and which are overstaffed at specific times.
Tuesday lunch at site 3 might be running at £18 per labour hour. That is your intervention point.
- Scheduled versus actual hours variance.
Track the gap between what was planned and what was worked.
If a site is consistently running 8 to 10% above schedule, the rota is wrong or the manager is not managing it.
- Labour cost by role category.
Separate kitchen, floor, and management into distinct lines.
Food cost and kitchen labour together give you prime cost, which is the number that tells you whether your core operation is profitable – not whether your revenue week was good.
- Fixed labour as a percentage of total labour.
In a well-run group, fixed labour (salaried management) should sit at a predictable ratio.
If it creeps up as you open more sites without a corresponding increase in the management structure it is funding, that is a warning sign about your overhead growth.
The broader point
Labour percentage is not a useless metric.
It is a useful summary number for investor reporting, lender conversations, and board-level review.
The mistake is using it as an operational tool.
Operational decisions – scheduling, headcount, rota design – need more granular data, faster.
By the time your monthly management accounts show a labour percentage that is 2% above plan, the weeks in question are already gone. The money is already spent.
The groups that run tight labour costs do not do it by watching a percentage.
They do it by knowing their revenue per labour hour by site by day, running daily comparisons of scheduled versus actual, and having managers who understand the numbers well enough to act on them in real time.
If your current reporting does not give you that, the issue is not your labour cost. It is your data.
Williams, Stanley & Co. work with hospitality groups across the UK to build financial reporting that is fast, granular, and built around how operators actually make decisions.
If your current management accounts are not giving you the operational visibility you need, talk to us.