Setting up clean books across several businesses

Almost nobody sets out to own several businesses. You buy a building and it needs its own entity for the mortgage. You start a second line of work and your lawyer tells you not to put it in the operating company. You hire one office manager for both and she starts paying both sets of bills out of whichever account has money in it. Four years later there are six legal entities, and the only person who understands how they relate is you.

That is the situation this article is about. Not accounting theory, and not a single small business with one bank account. The specific problem of an owner keeping several sets of books honest at once, with a small team and no controller.

What does bookkeeping for multiple businesses actually require?

It requires each entity to maintain a complete, self-contained ledger, plus a shared set of conventions that make those ledgers comparable and combinable.

Those are two separate obligations and most people only meet the first. Separate ledgers are a legal and tax requirement: each entity files its own return, holds its own assets, carries its own liabilities. Comparability is a management requirement. It is what lets you see that gross margin in one business is nine points below the other, and what lets you produce a group view without rebuilding it by hand every month.

Get the first right and the second wrong, and you are compliant and blind.

Start with one chart of accounts, copied deliberately

A group chart of accounts is a single numbered account list, maintained in one place, from which every entity's file is populated.

Adding an entity is a five-minute administrative act with a five-year consequence. The five minutes is creating a file in QuickBooks or Xero. The five years is that the file starts empty, whoever sets it up invents their own account names and coding habits, and nobody writes any of it down. Two entities later, "Contractors" means subcontracted labour in one file and the bookkeeper's fee in another.

So build the list once, for the most complicated entity you have, then copy it down. Entities that do not need an account leave it unused, which is fine and far better than the alternative. Resist letting each business customise. The customisation always feels reasonable in the moment and always costs you at consolidation.

Three rules that have earned their place:

• Number the accounts and keep the numbering identical everywhere. Same number, same meaning.
• Give intercompany balances dedicated accounts, one per counterparty, never mixed into general receivables or payables.
• Keep the list short. Departments, locations, and property-level detail belong in a dimension such as class or tracking category. Accounts multiply badly across entities; dimensions do not.

How should shared costs be allocated between entities?

Shared costs should be allocated by a written rule that is boring, stable, and applied every month whether or not anyone is looking.

Most groups have a handful of genuinely shared items: rent on a common office, one insurance policy, a software pool, one person doing admin for everyone. The temptation is to leave them where they were paid and sort it out at year end. That is how one entity ends up looking unprofitable and another better than it is.

Pick a basis, write it in one paragraph, stop revisiting it. Headcount for people costs, square footage for occupancy, revenue share for general overhead. The precision of the basis matters far less than the consistency of its application. A rough rule applied every month for three years produces usable trend data. A precise rule applied twice a year produces noise.

Intercompany: the entries most groups get wrong

An intercompany transaction is only correctly recorded when both sides are recorded, in the same period, at the same amount.

That sounds obvious and is violated constantly. Entity A pays a supplier invoice belonging to Entity B. Someone books it as an expense in A, because that is where the money left, and the corresponding receivable and payable never appear. Now A is overstated, B is understated, the two intercompany accounts disagree, and the group view will never tie.

The discipline is unglamorous. Every month, list every intercompany account across every entity and check that each pair nets to zero. If it does not, find it while you still remember what happened. Six months of unreconciled intercompany is no longer a bookkeeping task; it is an investigation, and catch-up bookkeeping is a separate project with its own sequence.

Banking hygiene, and the owner's card

Every entity needs its own bank account and card, and the owner's personal card needs to stop appearing in any of them.

I know the objection. Opening accounts is tedious and it is easier to put the group's subscriptions on one card. It is easier right up to the month you have to explain to a lender or a buyer why the operating company's statement contains expenses for three other businesses. Separate accounts turn a forensic exercise into a reconciliation.

If the owner's card is genuinely unavoidable, route those items through a single monthly reimbursement rather than dozens of individual entries, so the ledger shows one clean transaction with support behind it.

A worked example: three entities and a management company

Consider four entities. A is the original operating business. B is a second operating business started three years later. C holds the building both work out of. D is a management company employing the shared admin and finance staff.

Done well, this is not complicated. C charges A and B rent at a defined rate, invoiced monthly, both sides booked. D charges A and B a management fee based on headcount, also monthly, also both sides. Shared software sits in D and is covered by the fee rather than split line by line. Each entity shows its own revenue, its own direct costs, and one or two clean charges from its sisters. The group view is a roll-up with four elimination entries.

Done badly, the same structure produces rent charged in nine months of twelve, admin salaries stuck in A because that is where payroll was set up first, B looking artificially profitable, C showing a receivable A never recorded, and a year-end scramble in which the tax preparer makes allocation decisions on your behalf using judgement you never got to weigh in on.

Identical structure. The difference is entirely bookkeeping discipline.

What does a good monthly rhythm look like?

One calendar, one owner per task, the same sequence every month, all entities moving together.

The specific rhythm matters less than entities not falling out of step. Groups get into trouble when three entities are current and two are three months behind, because the group view is only as current as the worst entity. A slower cadence applied uniformly beats a fast cadence applied unevenly.

In practice: feeds reviewed weekly per entity, intercompany checked at every month end, and a fixed cut-off after which the period is closed and not reopened. The owner's guide to month-end close covers the sequencing in more detail.

Can QuickBooks or Xero do this?

QuickBooks Online and Xero are excellent single-entity ledgers and the right foundation for most groups, but neither was designed to manage a portfolio of entities as one thing.

Be specific about the gap. Both handle a single entity's transactions, feeds, reconciliation, and reporting very well, and both are affordable and well supported. What neither does natively is enforce one chart of accounts across many files, reconcile intercompany between files, eliminate those balances, or report the group as one set of numbers. That work happens outside the ledger, which for most groups means a spreadsheet.

That spreadsheet is where the risk concentrates. Panko's synthesis of audits of 88 operational spreadsheets found 94% contained at least one error, with an average cell error rate of 5.2%. Those are not exotic models; they are ordinary working files of exactly the kind used to combine entity trial balances every month. The broader consolidation problem is worth reading if that is where your group lives today.

What the data says about waiting

Groups usually discover their bookkeeping structure is wrong at the moment they add another entity, which is the worst possible time.

Intuit's Enterprise Technology Benchmark found 76% of firms operating multiple entities said their technology struggled when they added entities, 73% expected to outgrow their current stack within twelve months, and 64% said the close takes too long. APQC's Open Standards Benchmarking of roughly 2,300 organisations puts the median monthly close at 6.4 calendar days, the top quartile at 4.8, and the bottom quartile beyond 10. Multi-entity groups without shared conventions live in that bottom quartile more or less by construction.

Gartner's 2024 survey of 497 controllers and chief accounting officers adds the quieter cost: 18% of accountants report making financial errors at least daily, and 59% make several errors a month. Each of those is cheap to fix in the week it happens and expensive in the quarter it is discovered.

How cruisr approaches this

cruisr sits on top of the QuickBooks Online or Xero files you already have, one per entity, and treats the group as the unit of work rather than each file in isolation. AI keeps each entity's books current nightly, a human team reviews the exceptions rather than the whole population, and the group close package arrives by business day 3. There is no migration and the files stay in your name, which matters when you own the entities and not the tooling.

The part relevant here is less the automation than the enforcement: one account list, one calendar, intercompany checked every month across every entity, whether or not anyone remembers to ask.

Where to start

If you have four or more entities and a nagging feeling, do three things this month. Write down your chart of accounts and compare it against every file you have. Write your shared-cost allocation rule in one paragraph. List every intercompany account and check whether the pairs net to zero.

Whatever those exercises turn up is your actual project. Everything else can wait.

If you would like a second pair of eyes on it, cruisr runs a free close diagnostic on your own QuickBooks or Xero files and comes back within 48 hours with a 30-minute readout of what it found. You can get in touch here.