COGS as a percentage of revenue: what goes into it and what share is healthy

There is a question every seller eventually asks once the business is invoicing seriously: out of every peso that comes in, how much goes just to having the product there? That proportion — COGS as a percentage of revenue — is the metric that governs everything else. If it creeps up three points and you do not see it, those three points come straight out of your profit.

The problem is that almost nobody calculates it the same way. One seller counts only what the supplier charges. Another adds inbound freight. A third throws in the FBA fee because, they argue, without it the product never reaches the customer. All three have valid arguments, and all three report numbers that cannot be compared to each other, let alone to any industry benchmark.

In a marketplace business the confusion hurts more than in your own store, because a large share of the cost of serving the sale is charged by the channel, arrives blended into the settlement, and looks nothing like a supplier invoice. The FBA fee, the Full cost, the 3PL shipment and the unit that came back as a return are real costs per unit sold, but they live in four different reports.

This article separates the layers: what belongs in strict COGS, what is worth tracking alongside it even though it is also a variable cost, what proportions the available public measurements report — and which market they come from — and how to build your own baseline when no benchmark fits.

what COGS is and how it becomes a percentage

COGS stands for cost of goods sold. It is not what you bought, it is what you sold. If you bought 1,000 units at $260 and sold 400 during the month, your COGS for that month is $104,000, not $260,000. The other 600 units are inventory and live on your balance sheet, not your income statement. Confusing purchases with COGS is the most common accounting mistake among fast-growing sellers: it makes a heavy restocking month look like a losing month.

As a percentage, the formula is simple:

• COGS % = COGS for the period ÷ Revenue for the period × 100

Its relationship with gross margin is direct: if your COGS is 42% of revenue, your gross margin is 58%. Same information seen from two sides. Looking at it from the COGS side is worth the trouble because that is the actionable side — you cannot negotiate a gross margin, but you can negotiate a supplier cost.

the strict version of COGS and the extended one

Accounting COGS covers whatever is required to have the unit ready for sale:

• Supplier purchase price or contract manufacturing cost
• Inbound international and domestic freight, allocated per unit
• Duties, customs paperwork, broker fees and handling
• Primary packaging and labeling
• Direct labor, if you assemble or manufacture

What does not belong: marketplace commissions, advertising, storage, administrative staff, software. Those are selling and operating expenses, not product cost.

But marketplaces create a gray zone that matters enormously: the cost of serving the unit. The FBA fee, the Full rate, the shipping you absorb through a 3PL and the real cost of a return are variable costs that fire with every unit sold, exactly like the product itself. In accounting terms they are not COGS. Operationally they behave identically.

The practical answer is not to shove them into COGS and muddy the books, but to track them on a second line, always in view. Call it variable cost to serve, fulfillment cost, whatever you like, but measure it as a percentage of revenue next to COGS. Without that second line your COGS looks wonderfully healthy and your profit is nowhere to be found.

Glossary: inventory valuation, how much capital is tied up →

the four layers, with numbers

Take a month with $1,000,000 in revenue and pull it apart layer by layer. These figures are an illustrative example, not a market average.

• Product. Supplier cost of the units sold: $420,000, or 42% of revenue.
• Inbound logistics. Import freight, duties and handling allocated to what sold: $38,000, or 3.8%.
• Fulfillment. FBA and Full fees on the month’s orders: $95,000, or 9.5%. Plus your own shipping and 3PL: $30,000, or 3%.
• Returns. Units that came back and could not be resold, plus the logistics cost of bringing them in: $25,000, or 2.5%.

Strict COGS is the sum of the first two layers: $458,000, or 45.8% of revenue. With that figure your accounting gross margin is 54.2% and you feel comfortable. But the total variable cost of putting each unit in a customer’s hands is $608,000, or 60.8%. The difference is fifteen percentage points that exist, that get paid every month, and that simply do not appear in the narrow reading.

Those fifteen points are the distance between “my COGS is 46%, I am fine” and a profit that never materializes. And it is exactly where comparison against any benchmark breaks: if the reference figure measures product only and you measure product plus fulfillment, you are not comparing anything.

what proportions the public measurements report

It is worth stating up front where each figure comes from, because the universes are very different and none of them is the Mexican marketplace.

The most solid and verifiable reference is the industry margin database Aswath Damodaran maintains at NYU Stern, with data as of January 2026 covering publicly listed companies in the United States. The Retail (General) sector reports an aggregate gross margin of 33.18% across 23 companies, which implies COGS of roughly 66.8% of revenue. Retail (Special Lines), with 94 companies, reports 35.30% gross margin, that is around 64.7% COGS. The total market aggregate, almost 6,000 companies, sits at 37.76% gross margin. Translated: in listed US retail, two thirds of revenue goes to cost of merchandise, and the business defends itself on volume.

The other body of data comes from platforms aggregating Shopify stores. TrueProfit, across more than 5,000 analyzed stores, places typical gross margin between 55% and 65% — implying COGS of 35% to 45% — and describes the 60% to 70% gross margin band as the one that allows scaling. These are direct-to-consumer brands, mainly in the US market, selling their own product through their own store and paying no marketplace commission. Various DTC data aggregators repeat COGS ranges of 25% to 40%, but rarely publish methodology or sample size, so treat them as orientation rather than as a target.

And it is worth naming what does not exist: we found no public COGS or margin cut for Mexican marketplace sellers. The AMVO Estudio de Venta Online 2026 publishes an open version with market and shopper behavior data, but the operational performance cuts are reserved for its members. Polar Analytics publishes medians across more than 4,000 Shopify brands for conversion, ROAS, CAC and average order value, but not for COGS or gross margin. If someone hands you an exact percentage for “the average Mexican seller’s COGS,” it did not come from any published study.

why the healthy range depends on your model

With those two worlds on the table — 66% COGS in listed retail, 35% to 45% in DTC brands — the conclusion is not that one is right and the other wrong. It is that a healthy COGS depends on what you do with the product.

• If you resell third-party brands, your COGS will be high, probably above 55%, because you buy finished goods at wholesale. Your business is won on turnover, assortment and cost discipline, not on margin per unit.
• If you own the brand, your COGS should be substantially lower, but that cushion has to pay for everything brand building costs: photography, content, advertising, returns on a product the customer had never seen.
• If you import, your COGS moves with the exchange rate and with duties, and can shift two or three points between one purchase order and the next without you doing anything.

Category matters just as much. An electronics product competes in a market where price is anchored and COGS is structurally high; a private-label accessory plays a different game. Comparing yourself against an “ecommerce” average blends the two and tells you nothing.

how to set your own baseline

This is the part that can actually be done and that actually changes decisions. Five steps.

1. Freeze the definition and write it down. Two lines: what goes into strict COGS and what goes into variable cost to serve. Without that, your own time series is not comparable to itself.
2. Calculate on what sold, not on what you bought. You need unit cost per SKU and units sold in the period. If your cost changed between batches, pick a criterion — weighted average or FIFO — and stick to it.
3. Split it by channel. The same SKU carries a different cost to serve on Amazon with FBA than on MercadoLibre with Full. A consolidated COGS hides that one channel costs you more per unit.
4. Look at it per SKU and per family, not only globally. The business-level percentage is a volume-weighted average: your fastest movers dominate the number and hide the ones sitting at 75%.
5. Track the trend, not the level. That your COGS is 47% matters less than the fact that four months ago it was 43%. The level depends on your category; the trend depends on your own decisions.

With that baseline, supplier increases stop being a month-end surprise and become a data point you can bring to the negotiating table.

Glossary: 3PL, the logistics operator that stores and ships for you →

how COGS reads in iqseller

In iqseller COGS is a value you load per SKU, and from there the Profitability module puts it as the first line of the waterfall and subtracts the rest in the same place: commissions taken from the Amazon settlement and from MercadoLibre orders, FBA and Full fees, shipping, advertising, and the VAT and withholding breakdown. That lets you see strict COGS and variable cost to serve in a single view, which is exactly the separation the average spreadsheet never makes.

Because everything hangs off the Parent → Model → SKU tree, you can read the proportion at whatever level you need: the whole family when you negotiate with a supplier, the model when you compare variants, the SKU when you hunt for the one dragging the average. The Inventory module completes the picture through valuation: how much capital is parked and at what unit cost it is valued. And Alerts flag it when the proportion moves out of range, so a cost increase shows up the day it happens instead of six weeks later.

the number that actually moves decisions

COGS as a percentage of revenue is not for bragging or for comparing yourself against an infographic. It serves three concrete decisions: when to renegotiate with the supplier, when to raise price because cost moved, and when to stop restocking a product that no longer has room to pay for the channel’s structure.

For it to work it needs two things: a stable definition and a clean separation between the cost of the product and the cost of serving it. With those in place, the benchmark that matters stops being an industry that looks nothing like yours and becomes your own previous month.