Stop guessing whether your numbers are good. The 7 metrics every Singapore F&B operator should track weekly, anchored against the latest SingStat F&B Services Index.
Related:
Most Singapore F&B operators run on gut feel. Revenue felt soft last week. Labour cost looks high but no one is sure how high. Rent is what it is. This is how restaurants quietly bleed margin for months before anyone notices.
Benchmarking is the discipline of measuring against three reference points: your own past, the industry, and a structured framework of what healthy looks like. Done weekly, it takes 30 minutes. Done never, it takes the business.
This page is structured as a working reference, not a one-time read. Use it like this:
Reading time: 12 minutes end to end. Set-up time for the weekly cadence: 1 hour the first week, 30 minutes every week after.
Before benchmarking your own outlet, ground yourself in the wider Singapore F&B economy. These are the latest figures from the Singapore Department of Statistics Food & Beverage Services Index (FSI).
| Metric | Latest figure | Period |
|---|---|---|
| Total Singapore F&B sales value (monthly) | ~S$1.6 billion | March 2026 |
| Year on year sales growth | +2.3% | Mar 2026 vs Mar 2025 |
| Online sales share of total F&B | 20.6% | March 2026 |
| FSI base year | 2025 = 100 | Rebased Jan 2026 |
| Segment | YoY change | Reading |
|---|---|---|
| Food Caterers | +13.7% | Strongest growth; corporate and event demand returning |
| Fast Food Outlets | +4.8% | Steady growth; value-conscious consumers |
| Restaurants | +1.7% | Modest growth; competitive segment |
| Cafes | +1.1% | Soft growth; saturated market in many neighbourhoods |
| Food Courts & Other Eating Places | -1.5% | Declining; pressure from delivery and shifting consumer habits |
What this tells you: the Singapore F&B market is growing but unevenly. If your segment is shrinking (like Food Courts), beating last year requires winning share, not riding the wave. If your segment is growing (like Catering), flat results actually mean you are losing share. Context matters.
Online sales account for 20.6% of all F&B sales in Singapore as of March 2026, up from 20.0% the previous month. This is no longer a delivery surge or a pandemic hangover. It is a structural channel that any concept ignoring will progressively lose ground to competitors who plan around it.
Source: figures drawn from the Singapore Department of Statistics Monthly Retail Sales Index and Food & Beverage Services Index, March 2026, published 5 May 2026. We update this snapshot when new SingStat data is released.
You can measure almost anything in an F&B business. These are the seven metrics that drive every other number on your P&L. If you only ever track these, you will catch 90% of problems early.
What it is: COGS (food + beverage cost) plus total labour cost, expressed as a percentage of revenue.
Why it matters: the single most important number in F&B. Everything else (rent, marketing, profit) gets paid out of what is left after Prime Cost.
Track: weekly
What it is: the cost of food sold as a percentage of food revenue. Excludes beverage.
Why it matters: isolates kitchen performance from front of house and beverage. Sudden movement signals waste, theft, supplier price changes, or portion drift.
Track: weekly
What it is: total labour cost (wages + CPF + foreign worker levies + benefits) as a percentage of revenue.
Why it matters: labour is the lever you can adjust quickest. Track sales per labour hour (SPLH) alongside to make sure you are scheduling against demand.
Track: weekly
What it is: all-in rent (base + service charge + GST) as a percentage of revenue.
Why it matters: mostly fixed. If this number drifts above healthy ranges, your only real options are grow revenue, renegotiate, or exit. Worth knowing before you panic.
Track: monthly
What it is: total revenue divided by total covers (dine-in) or orders (delivery). Track them separately.
Why it matters: tells you whether menu engineering, upselling, and pricing strategy are working. Small AOV gains compound into significant margin.
Track: weekly
What it is: the number of covers served (dine-in) per day, or the number of times each table is occupied during a service period.
Why it matters: capacity utilisation. Flat AOV with falling covers is a demand problem. Strong covers with falling AOV is a pricing or product mix problem. Both together is the warning to act.
Track: daily, review weekly
What it is: cost of waste (spoilage, overproduction, plate waste, errors) as a percentage of revenue or food cost.
Why it matters: the silent margin killer. Most F&B businesses lose 4 to 10% of revenue to waste without realising. Measuring it makes it manageable.
Track: daily entries, review weekly
These are the working ranges most Singapore F&B operators target. They are not laws of physics. A premium concept will run different food cost than a fast casual, and a high-rent CBD unit needs different labour discipline than an industrial cloud kitchen. Use these as a starting point, then track your own trend over time.
| Metric | Healthy range | Watch zone | Red flag |
|---|---|---|---|
| Prime Cost % | 55% to 65% | 65% to 70% | Above 70% |
| Food Cost % | 28% to 35% | 35% to 40% | Above 40% |
| Labour Cost % | 25% to 35% | 35% to 40% | Above 40% |
| Rent % | Below 15% | 15% to 20% | Above 20% |
| Waste % | Below 4% | 4% to 8% | Above 8% |
| EBITDA margin | 10% to 18% | 5% to 10% | Below 5% |
Important context: these ranges are widely cited in F&B industry sources internationally and broadly applicable to Singapore operations, but they are not RAS-validated Singapore-specific benchmarks. They are a starting framework. RAS members can contribute to (and access) Singapore-specific benchmark data through our member survey programme.
Diagnose first: is it food, labour, or both? Run them separately. If food cost is the driver, audit recipes, portions, waste, and supplier prices. If labour is the driver, audit your schedule against actual demand and look for overstaffing on slow shifts.
Check the four usual suspects: (1) supplier price increases not passed through to menu pricing, (2) portion drift (cooks getting generous), (3) waste from overproduction or spoilage, (4) menu mix shifting to lower-margin items. Run an inventory variance to isolate.
Track Sales Per Labour Hour (SPLH) by daypart. You are almost certainly overstaffed at certain times. Build a demand curve from 8 weeks of data and reschedule against it. Trimming two hours from each weekday lunch can recover 3 to 5 percentage points.
The hardest one to fix. Three real options: (1) grow revenue (extended hours, delivery channel, catering), (2) renegotiate at renewal with evidence of comparable rents and your trading reality, (3) plan an exit. If your unit is structurally too small or too expensive for your concept, no amount of operational excellence will solve it.
Formulas are simple. Discipline is hard. These are the calculations to set up once and run weekly.
| Metric | Formula | Where to get the data |
|---|---|---|
| Prime Cost % | (COGS + Total Labour) ÷ Revenue × 100 | POS + payroll system |
| Food Cost % | Food COGS ÷ Food Revenue × 100 | Inventory + POS by category |
| Labour Cost % | Total Labour ÷ Revenue × 100 | Payroll (include CPF + levies) |
| Rent % | All-in Monthly Rent ÷ Monthly Revenue × 100 | Landlord invoice + P&L |
| AOV (dine-in) | Total Revenue ÷ Total Covers | POS |
| AOV (delivery) | Total Delivery Revenue ÷ Total Delivery Orders | POS + platform dashboards |
| Waste % | Cost of Wasted Items ÷ Food COGS × 100 | Waste log + inventory |
| SPLH | Revenue ÷ Total Labour Hours Worked | POS + timesheet |
| Line item | Monthly figure (S$) | % of revenue |
|---|---|---|
| Revenue | 120,000 | 100% |
| Food COGS | 38,400 | 32% |
| Beverage COGS | 4,800 | 4% |
| Total Labour (incl CPF + levies) | 36,000 | 30% |
| Prime Cost | 79,200 | 66% |
| Rent (all in) | 18,000 | 15% |
| Other operating costs | 14,400 | 12% |
| EBITDA | 8,400 | 7% |
Reading this example: Prime Cost is sitting at the top of the watch zone (66%). Food cost is on target. Labour is on the high side. EBITDA is in the watch zone at 7%. The lever with most upside here is labour, specifically scheduling discipline against actual demand. Pulling Prime Cost down by 2 points moves EBITDA from 7% to 9%, which is the difference between surviving and thriving.
Benchmarking that does not turn into action is just data. The point is the weekly habit that closes the loop from measurement to decision to outcome.
| Step | What you do | Time |
|---|---|---|
| 1 | Pull last week's numbers: Revenue, COGS, Labour, Covers | 10 min |
| 2 | Calculate Prime Cost %, Food Cost %, Labour Cost %, AOV | 5 min |
| 3 | Plot against the last 4 weeks (you are looking for trends, not single weeks) | 5 min |
| 4 | Identify the biggest gap or movement from healthy range | 5 min |
| 5 | Decide ONE action for the week ahead. Assign owner. Document. | 5 min |
The discipline that matters: one action per week. Not three. Not five. Operators who try to fix everything end up fixing nothing. Operators who fix one thing a week move 52 things in a year.
Keep it simple. A single row per week. Five to seven columns. The point is to make the trend visible at a glance, not to build a forensic accounting system.
| Week ending | Revenue | Prime Cost % | Food Cost % | Labour Cost % | AOV | One action this week |
|---|---|---|---|---|---|---|
| Wk 1 | S$28,400 | 64% | 32% | 32% | S$34 | Audit Sat dinner staffing |
| Wk 2 | S$29,100 | 62% | 31% | 31% | S$35 | Test set-menu upsell |
| Wk 3 | S$30,200 | 61% | 30% | 31% | S$36 | Review supplier prices |
| Wk 4 | S$31,500 | 60% | 30% | 30% | S$37 | Lock new supplier on dry goods |
The pattern above is what good benchmarking looks like in practice. Small, consistent movement in the right direction, week by week. No heroics. No firefighting. Just discipline.
Most operators ask the wrong question first. They ask "what is the industry average?" The better question is "what was I doing last month, and am I moving in the right direction?"
Your numbers vs your own past performance.
Your numbers vs the industry or comparable peers.
Your numbers vs what healthy F&B operations look like.
The order to use them: framework first to set rough targets. Internal weekly to drive operational decisions. External quarterly or annually for big strategic moves. Operators who only obsess over industry averages miss the trends in their own business that actually drive results.
A handful of situations where comparing yourself to industry data is genuinely useful:
Outside of these, focus on beating your own trend.
RAS is building the most accurate Singapore F&B benchmark dataset in the country, with figures contributed anonymously by RAS members. No more relying on US-sourced numbers that do not reflect Singapore reality.
Members get access to: Singapore F&B median Prime Cost % by segment (Restaurant, Cafe, Fast Food, Catering, Food Court), median Food and Labour Cost by concept type, median Rent % by location category (CBD, Orchard, Suburban Mall, HDB, Industrial), median AOV by cuisine type, EBITDA margin ranges by revenue band, and quarterly benchmark trend reports. All data contributed anonymously and validated by the RAS secretariat.
| Step | What happens |
|---|---|
| 1. Register interest | Tell us you want to contribute. We will keep you posted as the programme launches. |
| 2. Submit your numbers | Anonymous quarterly form. Takes 10 minutes if you have your P&L ready. |
| 3. Receive the report | Quarterly benchmark report showing how your numbers compare against RAS members in your segment. |
| 4. Use it | Plan around real Singapore numbers, not borrowed assumptions. |
Register your interest: contact the RAS secretariat at info@ras.org.sg with the subject line "Benchmark Survey" to be added to the contributor list. Members who participate get the full report. The more members contribute, the sharper the dataset.
Not yet a member? RAS membership unlocks the benchmark survey alongside the rest of the Members Vault. Learn more about RAS membership.
The most common ways operators sabotage their own benchmarking discipline. Recognise these before you fall into them.
Building a 30-column scorecard, getting overwhelmed in week 3, abandoning the whole thing. Start with three metrics. Add more once the routine is locked in.
One soft week is noise. Three soft weeks in a row is a signal. Wait for the trend before acting, otherwise you will whipsaw your team and your menu.
A 30-seat omakase will never have the same prime cost structure as a 120-seat zi char place. Comparing yourself to an irrelevant peer leads to wrong decisions. Compare against your own past and against concepts that look like yours.
Beautiful spreadsheet, perfect numbers, no decisions. The number is meaningless if it does not change what you do next week. Every weekly review must end with one decision.
If your segment (food courts, for example) is shrinking 1.5% year on year industry-wide, flat sales are actually a win. Always read your numbers against what the market is doing, not in isolation.
Asking your accountant for a monthly report is not benchmarking. The owner or GM has to be the one staring at the weekly numbers. Discipline does not delegate.
Most kitchens accept 6 to 10% waste as normal. It is not normal. It is just unmeasured. Start a daily waste log and you will be shocked. Then you will be motivated.
Use this checklist to audit your current benchmarking practice. Tick what you already do. The unticked items are your next 30 days of work.
Industry Benchmarks is the lens. The Pantry resources below are the tools to put it into practice.
Pattern to notice: benchmarks tell you what is wrong. The Pantry resources tell you how to fix it. Use them together.
