When should a self-employed owner hire an accountant?

Why the revenue-threshold answer is wrong

The most common version of this advice attaches a number — hire an accountant once you pass some level of turnover. It is memorable, easy to publish, and not much use, because turnover is a poor proxy for the thing that actually determines need.

Two businesses at identical revenue can be in completely different positions. One invoices a handful of clients monthly, has no staff and no stock, and could be managed on a spreadsheet by an organized owner. The other has employees, equipment, several revenue streams and inventory, and was already past manageable a year ago.

What determines need is complexity and friction — how many moving parts there are, how much time the admin takes, and whether decisions are waiting on information nobody has. A number cannot capture any of that.

The signals that actually indicate it is time

In practice the same handful of situations recur, and they are all about friction rather than scale.

The clearest is when you cannot answer a question about your own business without a weekend of work. If someone asks what the business made last quarter and the honest answer is that you would need to work it out, the record is not doing its job.

The second is a decision waiting on information. A hire, a large purchase, a loan application, a price change — each needs a reliable current position, and if getting one is a project, the decision gets delayed or made on impression.

The third is tax as an annual surprise. If the size of the bill is discovered rather than anticipated, nobody has been looking at it during the year, which is the only time anything could have been done.

The fourth is a structural change: taking on staff, changing entity, adding a revenue stream, working across a state boundary. Each introduces obligations that are cheap to handle in advance and awkward retrospectively.

And the fifth is simply time — when the admin has grown past the gaps available for it, and months are starting to slip.

• You cannot answer questions about the business without significant work
• A decision is waiting on numbers you do not have
• Tax is a surprise rather than something anticipated
• Something structural is changing — staff, entity, a new revenue stream
• The books are far enough behind that catching up feels like a project
• You are spending evenings on admin that the business should absorb

When you genuinely do not need one

This deserves saying plainly, because almost nobody selling accounting services says it.

A business with a small number of clients, no staff, no stock, a separate bank account and an owner who keeps on top of it may need an annual return and a conversation, and nothing more. Paying monthly for bookkeeping in that situation buys tidiness you already have.

The test is not whether help would be nice. It is whether the absence of help is currently costing you something — in time, in decision quality, in risk, or in money left on the table because nobody is looking. If the honest answer is no, the honest recommendation is to wait.

A provider who tells you otherwise before understanding your business is describing what they sell rather than what you need. It is a reasonable thing to test by asking directly whether they think you need the service.

Bookkeeping first, usually

Where help is needed, the first hire is usually bookkeeping rather than tax, and the sequencing has a logic.

Bookkeeping is the recurring work — the thing consuming evenings — and it produces the reliable record everything else depends on. Tax planning conversations built on unreliable books are speculation. Advisory conversations built on them are worse, because they attach confident conclusions to an unsound basis.

So the order that tends to work is: get the record right and current, then have the conversations that depend on it. Owners often approach it the other way round, hiring at tax time because that is when the pain is sharpest, and then discovering the preparer's first task is fixing a year of records.

The exception is a specific event with a deadline — a notice, an unfiled return, a transaction with a fixed date. Those set their own priority.

What it costs, and how to think about the cost

Pricing varies widely, and much of the variation is real rather than opportunistic: the same words cover very different amounts of work. Monthly bookkeeping might mean categorization from a bank feed, or it might mean reconciliation, a monthly close and someone who asks questions when something looks wrong. Those are different services at the same name.

The comparison that helps is not price against price but price against what specifically happens each period, and what happens when a transaction cannot be identified.

On whether it is worth it: the useful framing is not the fee against nothing, but the fee against what the alternative actually costs. That includes your own hours at whatever they are worth, the cost of a reconstruction at year end instead of a monthly close, and the harder-to-quantify cost of decisions made without information. For some businesses that comparison clearly favors hiring; for others it does not, and it should be allowed to say so.

What to ask before committing

The questions that separate providers are specific rather than general, and any provider worth using will answer them quickly.

• What exactly is done each period, and what is not included?
• Who reconciles the accounts, and how would I know it had been done?
• Will you tell me during the year if something looks wrong, or only at year end?
• What will I receive each month, and in what form?
• How do questions get handled — in both directions, and how quickly?
• What credentials do you hold, and what do they permit?
• If a tax authority writes to me, what is your role?
• What do you need from me, and by when?
• What would make you tell me I do not need this service?

A provider who answers these precisely is describing a process. General answers usually describe an intention, and the difference tends to surface at the least convenient moment.

Timing the first conversation

The worst time to start looking is the middle of filing season, when every practice is compressed into the same few weeks and onboarding competes with deadlines.

The best time is when nothing is urgent. A first conversation costs an hour, and having it before you need it means the relationship exists when something does happen. It also means the first year is not spent simultaneously building the picture and meeting a deadline.

If something structural is coming — a hire, a purchase, a change of entity — the conversation belongs before it rather than after. Most of the expensive mistakes in this area are sequencing mistakes rather than judgement ones.

What the first conversation should feel like

Mostly it should be questions about your business rather than a description of packages. A provider who cannot yet tell you what you need — because they have not looked — is behaving correctly.

You should come away with a clear sense of what is being proposed, what it excludes, what it costs and what happens next, and none of that should require chasing. Scope and price in writing before work begins is a reasonable expectation, not a demanding one.

You should not come away with a savings figure, a percentage or a threshold quoted before anyone reviewed your situation. Those numbers are produced by marketing rather than by your facts, and they are a reliable signal about what the rest of the relationship will be like.

You also do not need to tidy anything up first. Working out what state things are in is part of the job, and reorganizing beforehand usually creates work to undo.

If you leave a first meeting unsure what was actually being offered, that is information about how the rest of it would go.

The first year of working for yourself

One situation deserves singling out, because it produces more unpleasant surprises than any other: the year someone moves from employment into working for themselves.

In employment, tax is generally withheld as income is paid. Self-employment removes that mechanism without removing the underlying expectation that tax is paid as income is earned — the obligation moves to the individual, and nothing prompts it. Alongside that, self-employment carries tax considerations that employment does not, which is frequently the part nobody has mentioned.

The result is that someone who budgeted for income tax alone can find the total materially higher than expected, and can find it in April rather than in September when something could still be done about it.

Whether estimated payments apply to you, and in what amount, depends on your circumstances. This is the single clearest case for a conversation early in the first year rather than a return in the second.

Doing it yourself, well

If the conclusion is that you do not need help yet, it is worth doing the self-managed version properly rather than by default, because the habits you build now determine how expensive the eventual transition is.

Three things carry most of the weight. A separate business bank account, so the record is business-only by construction rather than by reconstruction. A monthly reconciliation, which is the check that reveals whether anything has been missed while the context is still fresh. And a consistent way of capturing receipts at the moment they exist rather than at year end.

A business doing those three things is in a genuinely good position, and when it eventually does need help, the first year with a provider is cheap because the picture already exists. A business doing none of them arrives with a reconstruction project attached.

Signs the current arrangement has stopped working

This question recurs for owners who already have a provider, and the signals are different from the ones above.

The common ones: you find out about problems after the year rather than during it; questions take a long time to be answered, or come back in language you do not follow; the same errors recur each year; your bookkeeper and your tax preparer have never spoken to each other; or you are receiving reports nobody has ever walked you through.

None of these individually mean you should move — some are worth raising first, and a good provider will fix them. What they do mean is that the arrangement is not delivering what it should, and that is worth naming rather than tolerating for another year.

Changing has a real cost: a first year with anyone new is more expensive in questions, because the picture has to be rebuilt. That is a reason to change deliberately when there is a genuine problem, and a reason not to change annually.

In short

Ignore revenue thresholds. Look at friction: whether you can answer questions about your own business, whether decisions are waiting on information, whether tax is a surprise, and whether something structural is changing.

If help is needed, bookkeeping usually comes first, because everything else depends on the record being right. Compare providers on what specifically happens each period rather than on price.

The first year of self-employment is the one exception worth treating as a default: even where ongoing help is not yet needed, a conversation in that year is usually worth having, because the mechanism that used to withhold tax for you has gone and nothing replaces it automatically.

And if the honest answer is that you do not need ongoing help yet, that is a legitimate answer — and it is worth asking a prospective accountant directly whether they think you do.

Whenever you do start looking, do it outside filing season. A first conversation costs an hour, and having it before you need it is the difference between choosing a provider and settling for whoever has capacity in March.