A restaurant chart of accounts should let the operator answer five specific questions in under a minute. If it can’t, it’s either too sparse or too cluttered. Both fail differently, and both get rebuilt at the first cleanup anyway.
Why the chart of accounts is the invisible foundation
The restaurant chart of accounts is the thing no operator wants to think about, and the thing that silently determines whether the P&L is useful or not. A COA that splits revenue the right way makes channel performance visible for free. A COA that lumps things together forces every report to be rebuilt from transaction-level data.
Every bookkeeping cleanup we scope starts here. Always. The first question we ask an operator with messy books isn’t “when was the last reconciliation.” It’s “can I see your chart of accounts.” The answer tells us most of what we need to know about where the books have been drifting.
The trap at setup: most restaurants start with the accounting software’s default template (“Sales,” “Cost of Goods Sold,” “Wages,” “Rent”) and never revisit it. Twelve months later, when the operator asks “how’s the delivery margin?” the answer is “pull the platform statements and build it manually.” The COA was the wrong tool for the question because the split was never there.
This guide is a practical starting point. It isn’t exhaustive. Exhaustive COAs have 300 accounts and nobody reads them. This is the structure that answers the five questions most hospitality operators actually ask about their business.
The five questions a good hospitality COA answers
- How did each revenue channel perform? Dine-in, takeaway, delivery by platform, catering, retail, private events
- What’s my food cost %, and is it moving? Food, beverage, and supplies separated
- What’s my labour cost %? Wages, benefits, casual labour, contractors, distinct.
- What are my real operating expenses? Rent, utilities, marketing, platform fees, software, not all lumped under “Expenses”.
- What’s my true EBITDA? Everything above operating income cleanly separated from depreciation, interest, and tax
A cafe coa that can’t answer these in under a minute of P&L reading needs restructuring. The sections below walk through each.
Revenue: split by channel, not lumped
The default sin of most hospitality chart of accounts setups: one “Food Sales” line that combines dine-in, takeaway, and every delivery platform. It balances. It also erases 80% of the signal.
A working revenue section:
4000 Revenue
4010 Food revenue: Dine-in
4020 Food revenue: Takeaway
4030 Food revenue: Delivery (Talabat)
4031 Food revenue: Delivery (Deliveroo)
4032 Food revenue: Delivery (Uber Eats)
4033 Food revenue: Delivery (Careem / other)
4040 Beverage revenue: Dine-in
4050 Beverage revenue: Takeaway
4060 Catering revenue
4070 Private events revenue
4080 Retail revenue (if applicable)
4090 Other operating income (cover charge, corkage)
4900 Service charge (if applicable, tracked separately)
4910 Tips collected (pass-through, not revenue)
Why this matters:
- Platform-level margin is visible. Talabat at 25% commission and Deliveroo at 30% look different in the P&L when they’re booked separately
- Channel mix shifts are visible. Dine-in dropping and delivery growing is a different business shape than dine-in staying flat while delivery grows. Only visible with the split.
- Tips in their own account (a pass-through liability, not revenue) prevents tips from inflating sales figures and distorting VAT / GST
COGS: food, beverage, supplies, separated
The second big split. Lumping all “cost of sales” into one bucket hides the two different margin stories most hospitality businesses have: food at 28-35% and beverage at 18-22%.
5000 Cost of Goods Sold
5010 Food: Meat and seafood
5020 Food: Produce
5030 Food: Dairy
5040 Food: Dry goods, pantry
5050 Food: Prepared / packaged
5100 Beverage: Coffee and tea
5110 Beverage: Soft drinks, juices
5120 Beverage: Alcohol (if applicable)
5200 Packaging and disposables
5300 Kitchen supplies (small ware, cleaning chemicals for BOH)
Most hospitality businesses don’t need finer detail than this at setup. Chefs’ meals / staff meals can go into a separate sub-account if they’re material; wastage can be a contra account to COGS if the process supports tracking.
Labour: the real second-largest hospitality cost bucket
Labour structure in the COA directly affects whether “labour %” means anything. A useful split:
6000 Labour
6010 Wages: Kitchen / BOH
6020 Wages: Service / FOH
6030 Wages: Management
6040 Casual / part-time wages (if tracked separately)
6100 Statutory benefits (super / WPS / mandatory benefits)
6200 Gratuity accrual (UAE) / long service leave accrual (AU)
6300 Payroll tax / social charges
6400 Contractors and outsourced labour
6500 Staff meals (if tracked as separate cost)
6600 Training and uniform costs
Keeping BOH and FOH wages split lets the operator see kitchen efficiency vs service efficiency as separate numbers. That matters when one is moving differently than the other, which happens more often than most operators expect.
Gratuity (UAE) and long service leave (AU) as separate accrual accounts is what prevents end-of-service payments from distorting a single month’s labour cost.
Operating expenses: what to split, what to combine
The trap here is opposite to revenue: too much detail. A COA with 40 expense accounts becomes a maintenance burden and the operator stops reading it. The right split is by decision category, not by invoice type.
7000 Operating Expenses
7100 Rent and lease
7200 Utilities (electricity, water, gas, internet)
7300 Marketing and advertising
7310 Platform fees (delivery platforms: commission as a % isn't a line item but the flat monthly / onboarding fees)
7400 Professional services (accounting, legal, consultancy)
7500 Software subscriptions (POS, accounting, scheduling, inventory)
7600 Bank and card processing fees
7700 Insurance
7800 Repairs and maintenance
7900 Cleaning and laundry (FOH)
7950 Supplies (office, printing, admin)
7990 Licensing and regulatory fees
Note: delivery platform commission is not in opex. It sits in its own expense section below COGS (or as a contra-revenue, depending on accounting convention), so commission can be analyzed alongside the platform revenue it relates to.
Platform commission: its own section
5500 Platform Commission
5510 Commission: Talabat
5511 Commission: Deliveroo
5512 Commission: Uber Eats
5513 Commission: Careem / other
This is the split that makes platform-level contribution margin possible. Revenue − COGS − Platform commission = what each platform actually contributes. Without this split, the platform analysis has to be rebuilt every month from scratch.
Balance sheet structure: the parts hospitality actually uses
The hospitality restaurant bookkeeping setup on the balance sheet side is often simpler than the P&L, but a few accounts need deliberate structure:
1000 Current Assets
1010 Cash on hand (by outlet, if multi-location)
1020 Bank: Main operating account
1100 Platform receivables: Talabat / Deliveroo / Uber Eats / etc. (clearing accounts)
1200 Inventory: Food and beverage
1300 Prepaid expenses (rent in advance, insurance)
2000 Current Liabilities
2010 Accounts payable: Suppliers
2100 Tips payable (from customer tips not yet paid out)
2200 VAT / GST payable (FTA / ATO)
2300 PAYG / payroll tax payable
2400 Super / WPS / gratuity accrual (non-current portion)
2500 Unearned revenue (catering deposits, event pre-payments)
Platform-receivable accounts, one per platform, as separate clearing accounts. That structure makes the weekly reconciliation between platform statements and bank payouts trivial. Lumping them into “Other debtors” hides what should be the most visible receivable the business has.
The COA rules that matter
- Name accounts for decisions, not for invoices. “Food cost: Produce” is a decision category. “ACME Vegetables Ltd invoices” is not.
- Avoid accounts with one transaction a year. If nothing lands in it, it’s noise. Collapse it into a broader category.
- Use sub-accounts sparingly. Three levels of depth are usually enough. Four is almost always a mistake.
- Keep account codes meaningful. Revenue accounts in 4000s, COGS in 5000s, labour in 6000s, opex in 7000s: the muscle-memory helps during monthly reviews.
- Review annually, not monthly. A COA change mid-year breaks comparability. Plan changes at financial year-end.
Common chart of accounts mistakes
- One “Sales” line. The single most limiting choice. Rebuild immediately if this is the current state.
- No platform commission split. Margin analysis becomes a manual exercise every time.
- Tips in revenue. Distorts sales, VAT / GST, and staff tax position.
- Lumped “Miscellaneous Expenses.” Becomes the dumping ground for everything the operator doesn’t feel like categorizing. Over a year, it’s usually the largest operating expense by accident.
- Too many expense accounts (50+). Nobody reads the detail; close variants get miscategorized; the P&L becomes unreadable.
- Multi-location with no location tagging. Outlet-level P&Ls require either location sub-accounts or tracking categories: not a single flat COA.
Starter checklist
- Revenue split by channel (dine-in, takeaway, each platform, catering, retail)
- Tips in a pass-through liability, not revenue
- Service charge in its own revenue account (VATable / GST-able)
- COGS split between food, beverage, packaging, supplies
- Labour split between BOH, FOH, management, and benefits / statutory
- Gratuity / long service leave as a separate accrual line
- Platform commission in its own expense section, split by platform
- Platform receivables as clearing accounts, one per platform
- Opex named by decision category, not by invoice type
- Location-level tracking configured if multi-outlet
Related resources
- Reading a restaurant P&L: the numbers that actually matter: the reports the COA feeds
- Xero vs QuickBooks for restaurants: the platform the COA sits on top of
- Month-end close checklist for hospitality businesses: the close rhythm that keeps the COA useful over time
Next step
If your current chart of accounts isn’t answering the five questions at the top of this guide, or you’re setting up books from scratch, the free books health check is the practical first step. We look at whether the COA fits the business or is working against it.
Last updated: April 2026.