Insights / Multi-unit and groups
Multi-unit restaurant accounting
How a restaurant group’s books actually fit together: separate entities per location, unit-level management reporting, and one consolidated set of numbers a bank will accept — plus a defensible way to allocate group overhead.
Published by Back of House Accounting · Principal: Douglas Carpenter, CPA, CFA · Last updated 27 September 2026
The three sets of numbers a restaurant group needs
Most groups that outgrow a single location end up with one set of books and a problem they cannot name. The books are usually organized around the legal entities, because that is what the tax returns need. What the operator actually wants to know — which location is carrying which — is nowhere on the page.
A group needs three views, and they are not the same document:
- Entity financials. One set per legal entity, because each entity files its own return and may have its own lender, landlord and ownership split. This is the statutory view.
- Unit-level management reporting. One profit and loss per location, on the same chart of accounts, in the same period calendar, with the same treatment of every line. This is the operating view, and it is the one that changes decisions.
- Consolidated financials. The group as a single economic unit, with transactions between the entities removed. This is the view a bank, an investor or a buyer asks for.
An entity is not always a location, and this is where a lot of group reporting quietly breaks. One entity can hold two locations. A location can be split across an operating entity and a real-estate entity. There is often a management company that employs the back-office staff and charges the locations a fee. If the reporting is built only on entities, none of the unit-level questions can be answered. Built only on units, the statutory and consolidated views cannot be produced without rework every year.
The fix is structural, not clever: one chart of accounts across every entity, a location dimension on every transaction, and one period calendar. Then all three views are the same data cut three ways. A group with lodging as well as restaurants runs the departmental version alongside it — the full hotel chart of accounts is published in the same way.
Which lines consolidate, which stay unit-level, and which eliminate
The single most useful page in a group reporting pack is the one that says where each number belongs. This is that page.
| Line | Unit-level P&L | Consolidated | Treatment and the usual mistake |
|---|---|---|---|
| Net sales | Yes, by location | Yes, summed | Report net of comps, discounts and voids. Gross sales less discounts is not the same figure as the deposit that reaches the bank, and reconciling to the deposit instead of to the POS sales report is the most common cause of a group’s sales never tying out. |
| Third-party delivery | Yes, by location | Yes | Record gross sales and the commission as a separate expense line. Booking only the net remittance understates sales, overstates margin and makes every percentage on the page wrong. |
| Cost of sales | Yes, by location | Yes | Needs a physical inventory count at period end per location. Purchases-as-cost-of-sales is fine within a period but makes unit comparison meaningless at the close. |
| Commissary or central kitchen transfers | Yes, both sides | Eliminated | The producing unit records a transfer out, the receiving unit a transfer in, at cost. Both are real for management purposes. Neither is a group sale, so the pair cancels on consolidation. |
| Direct labor | Yes, by location | Yes | Employees who work across two locations must be split by hours worked, not by where the paycheck was issued, or the better-performing unit absorbs the other’s labor. |
| Management or shared-service fee | Yes, as an expense | Eliminated | Expense at the location, revenue at the management company. Keep it visible at unit level so managers are measured on a cost they actually cause; eliminate it in the group column so the same dollar is not counted twice. |
| Occupancy | Yes, by location | Yes | If a related party owns the building, rent is an intercompany charge and eliminates on consolidation — but it must stay on the unit P&L at a market rate, or that location looks far more profitable than it is. |
| Intercompany loans and cash sweeps | Balance sheet only | Eliminated | Moving cash between entities is not revenue or expense anywhere. Groups that sweep cash nightly frequently have these sitting in an income account. |
| Group G&A and owner compensation | Allocated, shown separately | Yes | Show four-wall contribution first, then the allocation, then the result. Never blend the two into one number — see the allocation method below. |
| Franchise royalty and ad fund | Yes, by location | Yes | Accrue each period against the agreement’s stated basis. Only eliminates if the fund is an entity inside the same group, which for a franchisee it is not. |
| Gift card liability | Balance sheet, group-wide | Yes | Cards sold at one location and redeemed at another. Held as a group liability and settled between units, or the selling location keeps income it has not earned. |
Eliminations are the mechanical part of consolidation: a transaction between two entities inside the group is not income or expense for the group as a whole, so the matching pair is reversed in the consolidation column. What makes it hard in restaurant groups is not the arithmetic — it is that the transactions are often never recorded on both sides in the first place.
When separate entities have to be consolidated
Groups often ask whether they can simply present each entity separately and let the reader add them up. For a lender or an investor the answer is generally no, and the reason is a control test rather than a preference for tidier paperwork.
The Securities and Exchange Commission’s Regulation S-X states the principle plainly: consolidated financial statements “are more meaningful than separate financial statements and… they are usually necessary for a fair presentation when one entity directly or indirectly has a controlling financial interest in another entity.”
The practical translation for a restaurant group: where the same ownership controls the operating entities, the management company and often the property entity, the honest presentation is one consolidated set of statements with the intercompany traffic removed. Separate entity statements remain necessary for tax and for any lender secured against a single location — but they are an additional deliverable, not a substitute.
Ownership percentages differ from location to location in most growing groups, because a chef or an investor took a slice of one restaurant. That does not remove the consolidation requirement; it adds a non-controlling interest to account for. It is worth establishing which entities are in the consolidation group before the next set of statements is needed, not during a financing process.
This is a general description of reporting structure, not tax or legal advice for a specific group. Entity structure, ownership splits and lender requirements change the answer, and we are not your lawyers.
How to allocate group overhead without lying to yourself
A method you can run on your own numbers. It is also the step where most group reporting produces a confidently wrong answer.
Once there is a back office, an area manager, group insurance and software, somebody has to decide which location pays for them. Almost every group we see allocates that overhead as a percentage of sales, because it is the easiest defensible-sounding rule available. It is also the rule most likely to point at the wrong location.
Take a four-unit group. Four-wall contribution is calculated before any group overhead, and group overhead for the year is $520,000.
| Unit | Net sales | Four-wall contribution | Contribution % |
|---|---|---|---|
| Unit 1 | 2,400,000 | 384,000 | 16.0% |
| Unit 2 | 1,800,000 | 234,000 | 13.0% |
| Unit 3 | 1,250,000 | 100,000 | 8.0% |
| Unit 4 | 950,000 | 85,500 | 9.0% |
| Group | 6,400,000 | 803,500 | 12.6% |
The same overhead, allocated three ways
Group profit is $803,500 − $520,000 = $283,500 however it is allocated. Only the per-unit answer changes — and with it, which location the owner believes is the problem.
| Unit | A · by share of sales | B · flat per unit | C · split driver | C as % of sales |
|---|---|---|---|---|
| Unit 1 | 189,000 | 254,000 | 226,500 | 9.44% |
| Unit 2 | 87,750 | 104,000 | 97,125 | 5.40% |
| Unit 3 | (1,563) | (30,000) | (17,969) | −1.44% |
| Unit 4 | 8,313 | (44,500) | (22,156) | −2.33% |
| Group | 283,500 | 283,500 | 283,500 | 4.43% |
Method A, share of sales. Unit 4 takes 14.84% of the overhead because it produces 14.84% of sales, so it is charged $77,188 and reports a $8,313 profit. Unit 3 is charged $101,563 and lands just under water. The owner concludes Unit 3 is the weak location.
Method B, flat per unit. Each location takes $130,000. Unit 4 now loses $44,500 and is the worst performer in the group by a wide margin; Unit 3 loses $30,000. The owner concludes Unit 4 is the weak location.
Both cannot be right, and the difference is not a rounding argument — the ranking of the two worst locations inverts. That is the whole point: the allocation rule, not the operating performance, decided the answer.
Method C, and why it is the one to use
The driver should match how the cost actually behaves. Split group overhead into the part that is caused by having a site and the part that is genuinely group-wide:
- Per-site costs — $300,000 of the $520,000. Bookkeeping hours, area-manager coverage, per-location software seats, per-site insurance. A location’s books take roughly the same work at $950,000 of sales as at $2.4m. Allocate flat: $75,000 each.
- Genuine group costs — $220,000. Owner compensation, brand and marketing, legal, financing costs. These scale with the size of the business. Allocate by share of sales.
Unit 4 is charged $75,000 + $32,656 = $107,656 and loses $22,156. Unit 3 is charged $75,000 + $42,969 = $117,969 and loses $17,969. Both small units lose money, Unit 4 by more, and the two larger units are carrying them. That is the actionable truth, and neither single-driver method produced it.
The reason share-of-sales allocation misleads is specific and worth stating: most group overhead in a small restaurant group is a step cost per site, not a variable cost of revenue. Allocating it by revenue systematically flatters small locations and penalizes large ones. A group running this rule can keep a location open for years on the strength of an allocation choice.
Three rules make the method hold up: fix the split before you look at the answer, publish the drivers alongside the result so every manager can see the rule, and change the split only when the cost structure actually changes. An allocation nobody can see is an allocation nobody trusts.
Run the same discipline on the unit profit and loss statement first — an allocation laid over an unreliable four-wall number just distributes the error more confidently.
Exhibit A
Unit contribution summary
Four-wall contribution before group overhead, ranked. This is the page most operators have never seen.
| Unit | Net sales | Prime % | Occupancy % | Four-wall | Four-wall % | vs median |
|---|---|---|---|---|---|---|
| Downtown | 2,410,000 | 62.1% | 7.2% | 392,000 | 16.3% | +4.2 pts |
| Harbor | 1,980,000 | 64.8% | 8.1% | 271,000 | 13.7% | +1.6 pts |
| Midtown | 2,120,000 | 66.0% | 9.4% | 248,000 | 11.7% | median |
| Riverside | 1,640,000 | 70.4% | 11.8% | 96,000 | 5.9% | −5.8 pts |
| Group total | 8,150,000 | 65.6% | 8.9% | 1,007,000 | 12.4% | — |
What this exposes
Riverside generates 20% of group sales and 9.5% of group contribution. Its prime cost runs 8.3 points above Downtown’s, and its occupancy cost is 4.6 points higher — two different problems requiring two different responses. On a consolidated-only P&L, the group’s healthy 12.4% average conceals both entirely.
Exhibit B
Consolidation and eliminations
Separate entities per location, consolidated with intercompany eliminations and shared-service allocation shown explicitly rather than buried.
| Line | Units combined | ManageCo | Eliminations | Consolidated |
|---|---|---|---|---|
| Net sales | 8,150,000 | 489,000 | (489,000) | 8,150,000 |
| Cost of sales | 2,616,150 | — | — | 2,616,150 |
| Labor | 2,728,200 | 312,000 | — | 3,040,200 |
| Management fee to ManageCo | 489,000 | — | (489,000) | — |
| Occupancy | 725,350 | 42,000 | — | 767,350 |
| Other operating | 584,300 | 98,000 | — | 682,300 |
| EBITDA | 1,007,000 | 37,000 | — | 1,044,000 |
The management fee appears as an expense at unit level and revenue at ManageCo, then eliminates on consolidation. Shown this way, unit managers are still measured on a fee they pay, and ownership still sees a clean group number.
Exhibit C
Franchise accruals
Reconciled against the franchise agreement each period rather than estimated and trued up at year end.
| Obligation | Rate | Basis | Accrued | Paid | Balance |
|---|---|---|---|---|---|
| Royalty | 5.0% | 8,150,000 | 407,500 | 407,500 | — |
| National advertising fund | 2.0% | 8,150,000 | 163,000 | 163,000 | — |
| Local marketing minimum | 1.0% | 8,150,000 | 81,500 | 64,200 | 17,300 short |
| Technology fee | per unit | 4 units | 28,800 | 28,800 | — |
The local marketing shortfall is a contractual obligation, not a saving. Caught in period, it is a spending decision; caught at audit, it is a liability with a franchisor conversation attached.
The group reporting pack, and when it should arrive
Reporting that lands six weeks after period end is a history lesson. This is the cadence a multi-unit group should hold.
| When | Deliverable | What it answers |
|---|---|---|
| Weekly, day 2 | Flash by unit | Sales, prime cost and labor against plan for each location, while the week can still be changed. |
| Weekly, day 2 | Group cash position and 13-week forecast | Whether the group can cover payroll, rent and debt service, and which unit is consuming cash. |
| Period, by day 10 | Closed P&L and balance sheet, by unit and consolidated | Full four-wall contribution ranked, group overhead allocated on the published drivers, eliminations shown. |
| Period, by day 10 | Franchise and ad-fund reconciliation | Whether contractual obligations are accrued and met per unit. |
| Quarterly | Covenant and lender pack | Whether the group is inside its covenants, before the lender tells you it is not. |
| Annually | Entity statements and tax package | The statutory view for each entity, ownership splits included. |
We close the period profit and loss and balance sheet by day 10. The weekly flash exists because a period close, however fast, is still too late to change the period it reports on.
What changed for one multi-unit group
A multi-unit group came to us with books organized by entity and no unit-level reporting. Once every location was on one chart of accounts with a proper period calendar, we built the budget and the bank-ready consolidated financials the group had never been able to produce.
Those financials secured a credit line at a good rate. The group went from three locations to nine.
Anonymized at the client’s preference. We do not name clients.
Six mistakes that cost multi-unit groups the most
- Allocating group overhead by share of sales without asking whether the cost behaves that way. Demonstrated above: it can invert the ranking of your two weakest locations.
- Different charts of accounts per entity. Usually because each location was set up by whoever opened it. Every comparison between units is then an argument about definitions rather than performance.
- Calendar months across locations that run weekly operations. A month with five Fridays is not comparable to one with four, and in a group that error compounds across every unit at once.
- Intercompany transactions recorded on one side only. The management fee is expensed at the location and never booked as revenue at the management company, so nothing eliminates and the consolidation never balances.
- Related-party rent recorded below market, or not at all. The location looks strong until someone tries to sell it or borrow against it on those numbers.
- Consolidated-only reporting. The group average is healthy, so nobody looks harder. The unit contribution summary above exists precisely because a group total conceals the location that is funding the others.
Questions multi-unit operators ask
What is multi-unit restaurant accounting?
It is accounting built to report a restaurant group three ways at once: statutory financials for each legal entity, a profit and loss for each location on a common chart of accounts, and consolidated group financials with transactions between the entities eliminated. A single-location bookkeeping setup produces the first view only, which is why groups that scale on it lose the ability to see which location is performing.
Do I need separate books for each restaurant location?
You need separate reporting for each location, and separate books for each legal entity — which is not the same thing. If two locations sit in one entity, one set of books carries both, with a location dimension on every transaction so each can still be reported alone. Setting up disconnected accounting files per location is the more common and more expensive mistake: nothing consolidates without manual rework.
How do you consolidate financial statements for a restaurant group?
Put every entity on one chart of accounts and one period calendar. Combine like lines. Then reverse the transactions that occur between entities in the group — management and shared-service fees, commissary transfers, related-party rent, intercompany loans and cash movements — because a transaction inside the group is not income or expense for the group as a whole. Account for any ownership held by outside investors as a non-controlling interest. The eliminations should be shown as their own column, not netted invisibly.
What is a management company and why do restaurant groups use one?
A separate entity that employs the back-office and area-management staff and charges each location a fee for those services. It centralizes the group’s overhead in one place, keeps the cost visible on each unit’s profit and loss, and can simplify employment and benefits administration. For reporting it needs care: the fee is an expense at the location and revenue at the management company, and it must eliminate on consolidation or the group counts the same dollar twice.
How should group overhead be allocated across locations?
Split it by how the cost actually behaves. Costs caused by having a site — bookkeeping hours, area-manager coverage, per-location software and insurance — are step costs and should be allocated flat per unit. Genuinely group-wide costs such as owner compensation, brand marketing and financing scale with the business and can be allocated by share of sales. Allocating everything by share of sales flatters small locations and penalizes large ones; the worked example above shows it naming the wrong location as the group’s weakest.
When does a restaurant group need a controller or CFO rather than a bookkeeper?
The trigger is usually structural rather than a unit count. Once there are multiple entities to consolidate, an allocation policy to set and defend, lender covenants to track, or a financing or expansion decision in front of you, the work is no longer recording transactions — it is deciding how the group is measured. That is controller and CFO work. Many groups reach it at three or four locations.
How is franchise royalty accounting handled across multiple units?
Accrue each obligation every period against the basis the franchise agreement actually specifies — usually a percentage of net sales, with definitions that vary on discounts and delivery — rather than estimating and truing up at year end. Track royalty, advertising fund, any local marketing minimum and per-unit technology fees separately by location, and reconcile accrued against paid each period. A local marketing minimum that was under-spent is a contractual liability, not a saving. See franchise accounting for how we handle this.
Related: franchise and multi-unit accounting services · restaurant P&L statement · hospitality chart of accounts · CFO and controller services · restaurant cash flow