% COGS
What it means
% COGS is the share of a product's Gross Sales taken by what that product cost you to buy or make. Gross Sales is item price × quantity, counted before any discount comes off, so the denominator here is the list-price value of what sold — not what you banked.
Amazon carries a column with the same name and a different denominator: Net Sales, which already has refunds taken out of it. A smaller denominator against the same cost gives a bigger percentage, so the two numbers are not comparable and were never meant to be read side by side.
Show the math
Formula and a worked example
% COGS = COGS ÷ Gross Sales.
COGS is what the units cost you, taken from the cost price held against the variant. Gross Sales is item price × quantity, before discounts.
Worked example. A product sells 500 units at €40, so Gross Sales is €20,000. The units cost €14 each, so COGS is €7,000. % COGS = 7,000 ÷ 20,000 = 35%.
Now run the same 500 units through a 20% promotion. You take €16,000, the €7,000 of cost is unchanged, and cost is really eating 44% of what you got. This column still reads 35%, because its denominator never saw the discount. Read it beside % Product Discounts% Product DiscountsHow heavily a product is marked down.Under 8% is healthy or that gap stays invisible.
It answers the question
Of the list-price value of what you sold, how much was the product itself? It's the first cut taken out of a sale, and it sets the ceiling on every margin underneath it.
Why it matters
% COGS is the one cost you commit to before you sell anything. Everything else — ads, shipping, discounts — you decide week by week; this one was decided when you placed the purchase order.
It's also the fairest way to compare products of different sizes. A €12 item and a €300 item can't be ranked by profit in euros, but they can be ranked by the share of their price that goes on cost, and that ranking is what tells you which lines are worth pushing.
Because the denominator is pre-discount, treat it as a sourcing number rather than a trading one. It moves when your supplier price or your list price moves, and it sits perfectly still through a sale weekend that changed everything about what you actually earned.
What good looks like
The app publishes no bands for this column, and the honest target is your own. A category running at 70% product cost and one running at 25% are both real businesses, on completely different volumes. Track each product against its own trend and against the lines beside it in the table.
Two readings always mean something. A rise with your list price unchanged is a supplier or freight increase you haven't repriced for. And a % COGS holding steady while % Product Discounts% Product DiscountsHow heavily a product is marked down.Under 8% is healthy climbs means your true cost share is rising where this column can't see it.
The app does grade a related card, % COGS of Sales% COGS of SalesThe share of revenue lost to product cost.Under 35% is healthy, at under 35% healthy and over 55% needing work. That one divides by a different revenue figure, so borrow the shape of those bands, not the numbers.
How to improve it
| Lever | What you do | Expect | How long | Watch out for |
|---|---|---|---|---|
| Fast Fix the cost prices behind the column | Sync variant costs from Shopify so every line has a real cost against it | The number becomes true — often worse | Same week | A missing cost reads as 0% and flatters every profit figure below it, so the whole profit page gets uglier before it gets right. |
| Fast Reprice the lines carrying the highest share | Raise list price on the products sitting well above the rest of the table | % COGS falls without touching cost | 1–2 weeks | The ratio improves the day you change the price; demand answers weeks later. Watch Quantity for a full month before you call it a win. |
| Slow Buy deeper on the lines that actually turn | Move to a larger order quantity on your fastest sellers to earn a lower unit cost | % COGS down on those lines and it stays down | 1–2 quarters | Cash locked in stock, and markdown risk if demand turns before the depth sells through. The saving is real; the cash is gone. |
| Slow Shift the mix toward the cheaper-to-make lines | Put merchandising and ad budget behind the products with the lowest cost share | Blended % COGS falls with no product changing | 1–2 quarters | You're steering demand rather than fixing cost, so units can fall while the ratio improves. Check % New Quantity on the lines you starve before you cut them. |
Every lever costs something somewhere. The last column is the one to read twice.
Read it with
% COGS is what the product cost against its list price. What you actually banked is a different question, and the discount column is where the difference hides.
Margin from both ends
Cost share fell and you didn't pay it back in markdown. Rare, and worth understanding properly — it's usually one supplier change or one pricing decision, not a trend.
Given away at the till
You improved the cost side and handed some of it back in promotions. Deliberate during a launch or a clearance, expensive as a standing setting.
Costs rose, price held
Cost is taking more of the price and you're holding the line on markdowns. That's the right instinct, but it only holds while the units do — watch volume for a month.
Squeezed at both ends
Costs rose and you're discounting harder to keep the units moving. This is the one corner where doing nothing has a price attached.
They look like mirror images and aren't. This column divides by Gross Sales; Gross Margin divides by a revenue figure that has discounts, shipping, tax and refunds in it. When the two drift apart, the gap between them is everything that happens after the list price.
Cost share and volume share, on the same row. A high cost share on a product carrying 2% of your units is a rounding error; the same share on the product carrying 30% is where the money is going. Neither number ranks your problems without the other.
Common misreads
It means no cost price is synced against the variant. A blank cost reads as free and quietly inflates every profit number built on it.
It only ever covers product cost. Shipping, payment fees, advertising and the discounts you gave all sit outside this ratio, and any of them can turn a healthy cost share into a loss.
The denominator is pre-discount Gross Sales, so a promotion can't move this column by design. Margin fell; the number that would show it is the discount column next to it.
Also called
Cost of goods share · COGS ratio · cost of sales percentage
See yoursCost share per product on the Products table, in the same row as Gross Sales, Quantity and the discount columns.
Open Products →% COGS
What it means
% COGS on the Amazon settlements table is a SKU's cost of goods divided by its Net Sales. Net Sales is Revenue Sales minus Refunded Sales, so refunds come out of the denominator before the ratio is taken — this number already knows about your returns.
It's a settlement figure on both sides. Cost and sales are both read off the rows Amazon settled in the period, not off the orders your customers placed in it.
Show the math
Formula and a worked example
Net Sales is Revenue Sales minus Refunded Sales — FBA and MFN principal in, refunds back out. COGS is what the units on those rows cost you.
Worked example. A SKU settles €12,000 of Revenue Sales and €2,000 of Refunded Sales, so Net Sales is €10,000. Cost of goods is €4,000. % COGS = 4,000 ÷ 10,000 = 40%.
Hold that €4,000 of cost and take the refunds away: 4,000 ÷ 12,000 = 33%. Same product, same supplier, seven points of difference — all of it returns. That sensitivity is the thing to remember about this column.
It answers the question
Of the money Amazon actually settled for this SKU, how much was the product itself? It's the first cut, and it's taken before FeesFeesSum of commission, fulfilment, closing, storage, service, refund admin, shipping label, statutory deduction, and related fee fields for the row. and Tax, which sit in their own columns and are nowhere in this ratio.
Why it matters
Three things drive a settlement row and you control one of them. Net Sales moves with your price and your customers' returns, Fees are Amazon's to set, and cost of goods is yours. % COGS is where you watch the one you own.
It's also refund-sensitive in a way the Shopify column isn't. A SKU with returns creeping up sees its denominator shrink month after month while the cost per unit stays exactly where it was, so % COGS climbs for a reason that has nothing to do with sourcing. Refund RateRefund RateShare of ordered units that were refunded. is what separates the two.
What good looks like
No published bands for this column, and no useful universal number — a private-label SKU and a resold one carry completely different cost shares and both can work. Judge each SKU against its own trend and against the SKUs beside it on the same table.
Don't judge it against the Shopify column of the same name. That one divides by Gross Sales, before discounts and before refunds; this one divides by Net Sales, after refunds. On identical costs the Amazon figure reads higher, and the difference is denominators rather than performance.
The reading that always means something: % COGS rising while your supplier price is unchanged. That's the denominator moving — a price cut, or a run of returns — and Net SalesNet SalesSee the full entry. beside it will tell you which.
How to improve it
| Lever | What you do | Expect | How long | Watch out for |
|---|---|---|---|---|
| Fast Reprice where the denominator was cut | Check the SKUs whose cost share jumped for a price move rather than a cost move | % COGS back where it was, at the same volume | Same week | Repricing up costs units, and Buy Box % is worth watching closely the week after you do it. |
| Fast Cut the returns on the worst SKUs | Fix sizing, photos and description on the listings driving the highest refund rate | Less leaks out of Net Sales, so cost share falls with it | 2–6 weeks | An honest listing converts fewer of the browsers the old one over-promised to, so units can dip before the ratio improves. |
| Slow Take the cost out of the unit | Renegotiate or re-source the SKUs with the highest cost share and enough volume to justify it | % COGS falls and stays down | 1–2 quarters | A new supplier resets your quality and your lead times, and a stock-out during the switch costs you search rank as well as units. |
| Slow Retire the SKUs that never carry their cost | Delist or replace anything sitting at the top of the cost-share column for two quarters running | Blended % COGS improves with no negotiation at all | 1–2 quarters | Their units leave with them, and a thinner catalogue gives your advertising fewer places to work. |
Every lever costs something somewhere. The last column is the one to read twice.
Read it with
% COGS is the first cut out of a settled sale. It isn't the last one, and on Amazon the ones after it are large.
The ratio and its own denominator. Cost share up with Net Sales down is the denominator moving — a price cut or a bad month of refunds. Cost share up with Net Sales flat is a genuine cost increase. The percentage alone can't tell you which, and the two need completely different responses.
Your cost and Amazon's, side by side on the same row. A SKU can show a comfortable cost share and still lose money once fees are taken, because nothing in this ratio knows fees exist. Read them together before deciding a SKU earns its place.
Common misreads
Different denominators, so they can't match. Shopify divides by pre-discount Gross Sales, Amazon by Net Sales after refunds. Compare each to its own history instead.
Check the denominator first. Net Sales is after refunds, so a bad month of returns lifts this percentage with your unit cost untouched.
Fees and Tax sit in their own columns on the same row and never enter this calculation. What's left after product cost still has to pay both of them before any of it is yours.
Also called
%COGS · cost of goods share · COGS ratio
See yoursCost share per SKU on the settlements table, beside Net Sales, COGS and Fees for the same rows.
Open Amazon P&L →