With one bank account, every financial question requires analysis before you can answer it. Can I make payroll. Have I set enough aside for tax. Is that surplus actually surplus, or is it the deposit on a job I have not delivered yet.
The balance cannot tell you, because it is one number doing five jobs.
With accounts separated by purpose, most of those questions answer themselves. You look at the payroll account and you know. That is the entire argument, and it is a bigger deal in practice than it sounds in a sentence, because the questions you can answer instantly are the ones that stop occupying your attention.
It is the envelope system, done digitally. Nothing more sophisticated than that.
What this actually buys you
A real-time read without doing any work. Your account balances are your dashboard. No spreadsheet, no export, no mental arithmetic about what is committed.
Decisions made in advance. The best financial systems make decisions for you. When money lands and is immediately distributed by rule, you are not deciding in the moment whether this month can afford something. The rule already decided, at a time when you were thinking clearly rather than reacting.
Profit that exists. Profit treated as what is left over is not a plan, it is a hope, and there is never anything left over. Taking it off the top forces the business to operate on what remains, and if it cannot, that is information you needed.
A tax bill that is not a crisis. The most predictable financial shock in small business, and it flattens owners every year, because the money was spendable right up until it was not.
Less of your attention. This is the one owners underrate. Segmentation creates a kind of calm that is disproportionate to the mechanical effort involved. When money has a direction, you stop having a running argument with your bank balance.
The structure
Six accounts. You do not need all six on day one.
| Account | Purpose |
|---|---|
| Income | Everything lands here first. Nothing is ever paid from it |
| Payroll | Funded to a known weekly minimum, always |
| Operating | Rent, utilities, software, insurance, ordinary running costs |
| Taxes | Set aside off the top. Never borrowed from |
| Profit | Taken off the top. Accumulates |
| Reserve | Two to three months of operating expenses |
Income is a holding account, not a spending account. That distinction is what makes the system work. Money arrives, sits briefly and is distributed by rule. If you pay anything directly out of income, you have rebuilt the single-account problem with extra steps.
Payroll gets its own account and it matters most
In a service business where labor is the main overhead, this is the one that earns its keep immediately.
It means the only question that is ever genuinely urgent, can I make payroll, is answered by looking at one balance. Not by adding up receivables, subtracting commitments and estimating.
Work out your own number: the largest single week of payroll over your last quarter, plus a margin for the weeks you know run heavier. Once you know it, nothing else matters in that moment. That account needs that figure. Separating it puts the thing that could actually sink you in front of you, instead of buried inside one blended number.
If a different bill would sink you first, change the account and the label. A business with large recurring inventory purchases should ring-fence that instead. The principle is that the thing that will kill you first gets its own account.
Automate the transfers
Doing this by hand works and it relies on you remembering during a week when you are busy, which is exactly the week it matters.
Most modern business banks support rules that keep an account topped to a set balance or split incoming deposits by percentage. I use Mercury and have it move money daily so the payroll account always holds its minimum, and so profit and tax come off the top without a decision.
Two patterns worth setting up:
Split on arrival. A fixed percentage of every deposit goes straight to tax and profit before anything else sees it.
Top up to a floor. The payroll account is maintained at its minimum automatically, pulling from income or operating.
If your bank cannot do either, a standing weekly transfer on a fixed day is a workable substitute. The point is to remove the decision, not to have elegant automation.
What comes out first
Two questions decide this, and both are yours rather than mine.
What does tax actually cost you? It depends on your entity, your income and your jurisdiction. Your accountant can tell you in one email, and a guess here is expensive in both directions: too little and the bill is a crisis, too much and you have starved the business for no reason.
What are you trying to build? A business built to pay you now and a business built to be worth more later should not hold back the same amount. See is your margin healthy, or are you under-investing.
The structure is what matters, not the split. What you are building is the habit that profit and tax come off the top rather than out of what survives.
The consequence that catches people: some months you will look at the operating balance, see it flat, and conclude you are not making money. You are. The profit is in another account. If operating cash holds steady while the profit account grows, the business is performing well. The money is not gone, it moved somewhere you decided in advance it should go.
That reframe takes a few months to internalize and it is one of the more valuable ones available.
The reserve
Two to three months of operating expenses, and more if you are seasonal or growing quickly, because growth consumes cash faster than most owners expect.
As it grows, stop leaving all of it in cash. A checking account paying nothing is a slow loss. Move it somewhere liquid that earns something: treasury bills, a money market fund or similar. The requirement is liquidity, not zero yield. You need to be able to reach it in days, not the same afternoon.
Worth distinguishing two things owners conflate:
Cash on hand is what is actually available across your accounts. Cash reserve is money set aside for a defined purpose: a vehicle, a build-out, a known seasonal gap.
They are different numbers, and treating earmarked money as available reserve makes you feel safer than you are.
The reserve is also the thing that quietly improves your judgment everywhere else. Reserves solve a surprising number of problems that present as operational. Waiting on a payment is only expensive if the delay means a hard conversation about payroll. Fix the reserve and the same late payment stops mattering, which means you stop building rigid processes around your own anxiety.
Start with three
The full structure is a lot to adopt at once, and more accounts is more administrative overhead. If you are not going to maintain the transfers, six accounts will make things worse rather than better.
Income, payroll, taxes. Those three deliver most of the benefit. Add operating, profit and reserve once the first three are habitual, usually a month or two.
The tradeoff, stated plainly: this is real setup work and ongoing discipline for a benefit that is mostly cognitive. You will not make more money next quarter because of it. You will make better decisions, faster and with less of your attention consumed, and over a year that compounds into something measurable.
The same logic outside the business
Worth a short note, because the principle does not stop at the company.
The same segmentation works for personal finances, and for the same reason: when every dollar has a named destination, you stop deciding and start executing. Taxes, investing, property, giving, ordinary living costs and discretionary spending can each have a home.
Underneath both is a view about money worth stating directly. Money is a fantastic employee and a terrible owner. Directed deliberately, it builds what you point it at. Served anxiously, it sets the agenda.
And having it amplifies whoever you already are. It does not make people generous or greedy. It magnifies what is there and removes whatever was previously stopping you. Which is an argument for building the discipline before the abundance rather than after, because the system you set up while the numbers are small is the one you will still be running when they are not.
Do this
- Open three accounts: income, payroll, taxes
- Set the payroll minimum to one week’s payroll, plus a margin
- Automate a split on arrival for tax, and a top-up to the payroll floor
- Route every incoming payment to income. Pay nothing from it
- Run it for a month, then add operating, profit and reserve
- Build the reserve to two months of operating expenses before you fund anything else
13-Week Cash Flow Forecast template → The forecast assumes this account structure. Together they are the system.
Frequently asked
Will my bank charge for multiple accounts? Some do. Many business banks offer several sub-accounts at no cost, and it is worth switching for. If yours charges real money per account, that is a reason to change bank rather than to abandon the structure.
Does this affect my bookkeeping? It usually simplifies it, because transactions arrive pre-categorised by which account they touched. Tell your bookkeeper before you start so the chart of accounts matches.
What if a week is short and payroll needs money from the tax account? Then you have found a real problem and the system did its job by surfacing it. Borrowing from tax is a decision, not an accident. Record it, and put it back. If it happens twice, the issue is not the account structure, it is pricing, labor cost or timing.
Is this just Profit First? It draws on the same idea. The adaptation here weights payroll far more heavily, because when most of your money goes to wages, payroll is the bill that ends the business fastest.
How does this work with a line of credit? Keep it separate and treat a draw as an event, not as a balance. A credit line blended into your operating account restores exactly the ambiguity the structure removes.
What if my income is irregular? The structure helps more, not less. Percentage-based splits on arrival handle variable income naturally, and the reserve is doing the smoothing that a predictable salary would otherwise do.
This and the thirteen-week forecast are usually the first two things I set up with an owner, because almost every other decision reads more clearly afterwards. Here is how I work.