The Amazon seller accounting spreadsheet: when it works, and when it quietly stops
Almost every Amazon seller starts with a spreadsheet, and for a while it’s the right call — no subscription, no setup, and you already know how to use it. The question isn’t whether a spreadsheet can track an Amazon business. It’s at what point it stops telling you the truth without telling you it’s stopped.
Where a spreadsheet genuinely holds up
If you’re pre-revenue or doing a handful of sales a month, a spreadsheet with a few columns — units bought, units sold, fees, revenue — is honestly fine. You can eyeball every row. Nothing is complicated enough yet to hide a mistake.
Where it starts to break
Inventory that moves through more states than a single row can hold. A unit gets ordered, prepped, shipped, received by Amazon, damaged, partially credited by the supplier, reimbursed at a different amount by Amazon, or returned and re-sold. A spreadsheet row can hold a quantity; it can’t hold that whole history without you rebuilding the logic yourself, every time, by hand.
Landed cost per SKU. Unit price is easy. Freight, duty, and prep fees allocated across units in a shipment is a real calculation, and if you’re not doing it, your “cost” column is understating what each unit actually cost you — which means every profit number downstream of it is wrong in the same direction.
Settlement deposits that aren’t one transaction. Amazon pays you every two weeks, net of dozens of fee types, refunds, and adjustments, in a single deposit. A spreadsheet built around “money in, money out” either records the whole deposit as revenue (overstating it) or requires you to manually decompose every settlement report by hand, every two weeks, for as long as you sell.
The 1099-K, at year end. Box 1a on the 1099-K is gross, unadjusted payment volume — no netting for refunds or fees. If your spreadsheet’s “revenue” column doesn’t match that number, and it usually won’t, you need a real bridge between the two, and building that bridge in a spreadsheet from scratch each January is exactly the kind of one-off, error-prone work that a formula typo turns into a wrong number nobody catches.
The point you don’t see coming. None of the above breaks loudly. A spreadsheet doesn’t throw an error when a formula is subtly wrong or a row got overwritten — it just keeps calculating, confidently, on bad data. The first sign is usually a number that doesn’t feel right months later, by which point untangling what went wrong costs more time than building it correctly would have.
What replaces it
Real double-entry accounting doesn’t ask you to model any of this by hand. Inventory moves as an asset through every state above and lands in cost of goods sold only when a unit actually sells. Every settlement deposit is decomposed into its real components automatically. The 1099-K bridges to your books by formula, not by a spreadsheet you rebuild every January. None of this requires giving up the parts of a spreadsheet you actually liked — it requires the parts you were doing by hand to stop being your job.