NetNet vs a Profit Spreadsheet

Most stores calculate profit in a spreadsheet, and most of them should keep doing it for longer than software vendors suggest. This page is about where the line actually falls, not about why a subscription is obviously better.

What a spreadsheet does well

It costs nothing. It bends to any business model, however unusual — bundles, subscriptions, wholesale alongside retail, a channel nobody has an integration for. And whoever built it understands every formula in it.

That last property is worth dwelling on. A number you constructed yourself is a number you can defend, debug and adjust. Software output has to be taken on trust until you have learned where each figure comes from, and plenty of merchants have replaced a spreadsheet they understood with a dashboard they did not, then quietly stopped looking at either.

Building the calculation by hand once is also the best way to learn what actually drives your margin. We would rather someone did that first than bought an app to skip it.

Where it stops working

Two failures, and neither is about arithmetic.

Freshness. A spreadsheet is exactly as current as the last time someone rebuilt it, which for most stores is monthly and in practice means the working number is two weeks old on average. Every decision taken in between is taken against a stale figure. This is not a precision problem — the spreadsheet may be perfectly accurate about a period that has already ended.

Netted costs. A manual file is assembled from things a person can see: sales in Shopify, invoices in an inbox, spend in an ad account. The costs it misses share one property — they are deducted rather than billed. Payment fee tax, foreign exchange spread on payouts, carrier weight adjustments, fees retained on refunded orders. None produces a document anyone has to act on, so none prompts a row. The omission is structural, not careless, and it runs in one direction: everything missing is a cost, so the spreadsheet is systematically optimistic.

The threshold that actually matters

Not revenue. Not order count. Two questions:

How many separate systems do you reconcile? One carrier and one gateway is manageable by hand. Add a second carrier, a 3PL, a second payment method and an Amazon channel, and the monthly rebuild grows past the time anyone will reliably give it.

How often do closed months reopen? Every late refund and carrier adjustment either gets attributed back — reopening a month — or booked to the wrong period, which quietly corrupts per-product and per-channel figures. Spreadsheets handle this badly, and the usual resolution is to stop attributing properly.

The honest signal is simpler than either: the month it does not get rebuilt. Not because anyone decided to stop, but because it took four hours and something else was on fire. That is the point at which a spreadsheet has stopped being the tool and started being an intention.

What you give up

Flexibility, mostly. Fixed reports answer the questions they were built for. A spreadsheet answers whatever you can express in a formula, including questions specific to your business that no product will anticipate.

If your model is unusual, that flexibility may matter more than freshness, and the right answer is to keep the spreadsheet.

Running both for a month

If you do switch, run both in parallel for one period and reconcile them line by line.

The differences are the interesting part. Either the spreadsheet was missing costs — usually netted ones, usually two to four percent of net sales — or the app is configured wrongly, most often an uncosted variant or a shipping rule that does not match reality.

Either finding is worth the afternoon. And having done the reconciliation once, you will know where the software’s numbers come from, which is the trust problem solved rather than deferred.

Keep the spreadsheet afterwards, too. It costs nothing to leave in place, and it remains the fastest way to model a question no product anticipated — a new channel, an unusual bundle, a what-if on supplier pricing.