A spreadsheet is a legal set of books. The IRS says so without hedging: “You may choose any recordkeeping system suited to your business that clearly shows your income and expenses” (IRS: Recordkeeping). No software is prescribed, no format is blessed, and nobody is going to disallow a deduction because the ledger was a grid instead of a subscription.
That settles the legal question and leaves the useful one. If the rules do not require the app, what does it actually do that the grid does not?
Disclaimer: This article explains published IRS recordkeeping guidance and is not tax advice or financial advice. Consult a tax professional about your own situation.
The thing the IRS values is timing
Buried in the travel and expense rules is a sentence that decides the whole comparison: “a record made at or near the time of the expense has more value than a record made later” (IRS, Publication 463). The same publication asks for “timely kept records” and tells you to “record your expenses at or near the time you have them” (IRS, Publication 463).
Notice what that is not. It is not a rule about software. It is a rule about when the writing happens, and it is the one dimension on which a spreadsheet and an app genuinely differ.
A bookkeeping app records at the moment the money moves, because the bank feed does it whether you are paying attention or not. A spreadsheet records when you sit down to fill it in. For a disciplined person those are the same thing. For everyone else they are a month apart, and the gap is not a filing inconvenience — it is the difference between evidence and recollection.
So the real question is not which tool is better. It is whether you will open the spreadsheet on a Tuesday in March when nothing is due. Most people know their own answer to that and argue about features instead.
Four things that end the argument regardless
Discipline settles the ordinary case. Four situations settle it for you, and any one of them is enough.
You are depreciating something. The IRS wants asset records showing “When and how you acquired the asset. Purchase price. Cost of any improvements. Section 179 deduction taken. Deductions taken for depreciation” (IRS, Publication 583). That is not a row in this year’s sheet. It is a running position that has to survive every year until you sell the thing, and a spreadsheet carries it only as long as you keep copying it forward correctly.
You have money owed to you. A ledger of what happened is a different object from a list of what has not happened yet. Unpaid invoices need chasing on a schedule, and a system that only records completed transactions gives you no view of the ones that never completed.
Your transaction count is past what you will reconcile by hand. The test is not the number itself, it is whether you will match every line against the bank statement each month. If you know you will skip it, the count is already too high.
Anything at all involves inventory. Goods bought for resale bring in a second running position, and running positions are the specific thing spreadsheets lose.
Notice that none of these is a profit level. Profit is the number people reach for because it is the one they know, and it correlates only loosely. A consultant clearing well into six figures with fourteen transactions a year and no assets can run a spreadsheet indefinitely. A much smaller operation with a van on a depreciation schedule and forty unpaid invoices cannot.
Where the spreadsheet genuinely wins
It is free, it is yours, it opens in twenty years, and nobody can raise its price or discontinue it. Those are not small things, and the advice industry underrates them because nobody earns a commission on a grid.
A spreadsheet is also the better teacher. Building your own categories forces you to understand what a category is for, which is knowledge an app quietly does on your behalf and then hides. Plenty of people should spend their first year in a spreadsheet precisely so the second year in an app makes sense.
And there is a shape of business the grid simply fits: few clients, few categories, no assets, no inventory, everything paid on receipt. That business does not have a bookkeeping problem to solve.
The switch that is not a switch
If you do move, the useful thing to know is that you are not abandoning the spreadsheet so much as demoting it. The app becomes the record of what happened. The spreadsheet stays for the thinking — pricing, capacity, what a client is actually worth after the hours are counted. Those are the questions a bookkeeping app is not built to answer and never will be.
Keep the records for as long as the return they support is open, which “generally” means “until the period of limitations for that return runs out” (IRS, Publication 583). Whatever holds them has to still open then, which is an argument for exporting from any app you use, on a schedule, in a format that outlives the subscription.
If you sign up through the link below, we may earn a commission. It costs you nothing extra, and it does not change what this page says about the number or the rule.
If one of the four triggers above describes you, the job is to get transactions recorded as they happen rather than remembered in April. [Product] pulls from the account and categorises as the money moves, which is the “at or near the time” standard the IRS actually values. It earns its keep once you are depreciating an asset, carrying unpaid invoices, or past the transaction count you will reconcile by hand, and it is hard to justify below that. [See it →]
Two neighbouring decisions are worth settling in the same sitting. The separate business account is what makes any of this legible in the first place, and it costs nothing. The 1099 tax estimator turns the resulting numbers into the quarterly payment, which is the only reason most people started keeping books at all.
The spreadsheet did not stop working because it was a spreadsheet. It stopped working on the month you filled it in from memory, and you already know which month that was.
