Every accounting file starts with a default chart of accounts. It is generic by necessity — it has to be plausible for a law firm and a food truck. The result is that it is genuinely useful to neither.
Most businesses never change it. Then, two years later, they wonder why their reports do not tell them anything.
The purpose, stated plainly
A chart of accounts exists to answer questions. Not to satisfy an accountant, not to be comprehensive — to answer the specific questions you ask about your business.
Which means the design process starts with the questions, not the accounts. Write down the five things you most want to know each month. Then build the structure that produces them.
A contractor wants labour, materials, and subcontractors separated, because that is how a job is priced. A restaurant wants food and beverage split, because they carry different margins and drift independently. An online seller wants platform fees visible as their own line, because it is one of their largest costs and it changes without notice.
None of those distinctions exist in the default list.
Cost of goods sold versus operating expenses
This is the split that produces gross margin, and getting it wrong makes the single most useful number in your P&L meaningless.
Cost of goods sold is what scales with delivery — materials, direct labour, subcontractors, merchant fees, shipping. Operating expenses exist whether you sell anything or not — rent, insurance, software, admin salaries.
Put direct labour into operating expenses and gross margin looks wonderful while the business loses money. Put rent into cost of goods sold and margin collapses for no reason. The test is simple: if revenue doubled next month, would this cost roughly double? Then it is a direct cost.
How detailed
The most common mistake is too much detail, not too little.
Forty expense accounts feels thorough. In practice it means whoever is coding transactions has to make forty-way decisions, which produces inconsistency — and inconsistent coding is worse than coarse coding, because at least coarse coding is reliably coarse.
The rule that works: create an account when you would actually make a decision differently based on seeing it separately. You would act on knowing fuel cost. You would probably not act on knowing paper clips separately from printer ink. That is one Office Supplies account, not two.
If you want detail without account sprawl, that is what classes, jobs, and locations are for — a second dimension that lets you slice by segment without multiplying the account list.
Consistency beats correctness
An imperfect structure applied consistently for three years is more valuable than a perfect one adopted in June.
Comparability is the point. If a cost moved between accounts mid-year, every year-over-year comparison through that boundary is broken, and nobody remembers why the numbers jumped. When restructuring is genuinely needed, it belongs at a year boundary, with a note recording what changed.
Signs yours is not working
- A single account holds more than about 15% of total expenses — usually a catch-all doing too much work.
- Accounts with near-identical names, created because nobody could find the existing one.
- Accounts with no activity for two years.
- You routinely export to a spreadsheet to answer a basic question, because the report cannot produce it directly.
That last one is the clearest signal. If the answer always requires a spreadsheet, the structure is not doing its job.
Getting this right at the start is most of what setup and onboarding is, and rebuilding it is usually the first step of a cleanup — because there is little point categorising two years of transactions into a structure that cannot answer anything.
Behind on Your Books?
We handle your bookkeeping end-to-end — categorization, reconciliation, month-end close, and clean financial statements, so your books stay current and CPA-ready.