Prime Day's Over.
Your Margin Report Isn't.

The event ends. The settlement doesn't

A discount is instant. Its cost is not. You gave up margin dollars in a weekend; you'll spend a quarter finding out how many.

The marketplace discount cycle — Prime Day, then Labor Day, then the Black Friday run, then a “New Year” clearance — trains two things at once. It trains inventory to move. It also trains your customer to wait, which quietly repriced every full-price week between the events.

Contribution margin (what's left from a sale after the variable cost of making and delivering that unit) is the number that actually pays your rent. Public DTC apparel brands run a median gross margin around 55–57% (Eightx, SEC 10-K data) — and still go underwater at the contribution line once the discount and the fulfillment stack eat the rest.

“The revenue spike is a number you feel on the day. The margin cost is a number you read a quarter later — and by then the next discount window is already open.”

The make-vs-discount math founders skip

Here's the math that gets skipped, because it's uncomfortable and it isn't on the sales dashboard. Take a $50 tee at a 40% gross margin — $30 to make and land, $20 of margin. Watch what each discount tier does to the units you now have to sell just to keep the same profit dollars (Phoenix Strategy Group).

A 30% cut doesn't cost you 30%. It quarters your margin per unit, so you need four times the volume to stand still. Prime Day gives you a volume bump — rarely a 4× one. The gap is your loss, and it posts in July.

“A 30% discount doesn't cost 30%. It quarters your per-unit margin — and the marketplace fee comes out of what's left.”

The marketplace tax nobody re-priced

If any of that volume ran through Amazon, a second layer came out before you saw a cent. Apparel referral fees are tiered — roughly 5% under $15, 10% from $15–$20, and about 17% above $20 (EcomCalc), and they're frozen at those levels through 2026.

Fulfillment moved, though. As of April 17, 2026, Amazon added a 3.5% surcharge on all US FBA fulfillment fees, on top of a January fee revision (Amazon Selling Partners). Most founders re-priced their retail for the discount and never re-priced for the fee. On that same $50 tee, ~17% referral plus fulfillment plus the discount can leave a contribution margin you'd never have greenlit on paper.

Why “just move units” is the trap

The instinct after a soft margin report is to run another promo and make it up on volume. In apparel, three forces make that the most expensive instinct you have.

• Returns. Apparel returns run 25–40% of gross sales, and a 30% return rate can raise your realized customer acquisition cost by around 43% on the customers who keep the product (Luca). A discount-driven buyer returns more, not less.
• Repeat rate. Apparel has the lowest repeat-purchase rate in DTC — under 10% (Eightx). A discount buyer rarely comes back at full price to fix the economics.
• CAC. Apparel customer acquisition costs sit near $90 (Eightx). Spend that to acquire a one-time, deep-discount, high-return buyer and the unit was underwater before it shipped.

The counterexample — when the discount is the right call. Sometimes it is. Clearing end-of-life inventory, funding a genuine first-order acquisition play with a real repeat engine behind it, or avoiding a carrying cost that exceeds the margin you'd give up — those make a deep cut the disciplined move, not the desperate one. Markdown optimization (right depth, right timing) is shown to lift margin rate by 4–8 points versus reflexive, across-the-board cuts (Onebeat). The problem is never discounting. It's discounting without the math.

Building margin back in — before the next drop

Here's the lever the discount conversation skips entirely: the make side. A discount subtracts margin per unit and trains the customer to wait. A lower landed cost adds margin per unit — at full price, at every price, on every future drop, with no behavior to un-train. They point in opposite directions.

Shave that $50 tee's landed cost from $30 to $27 — through better fabric sourcing, a tighter tech pack, smarter trims, consolidated production — and you've added $3 of margin per unit at full price. That's more durable profit than a 15%-off weekend delivers, and the customer never learns to hold out for it. It's the difference between the real cost of the wrong clothing manufacturer and a make partner who builds margin headroom in at the tech-pack stage.

Do it before the sample is approved, not after the settlement lands. And read it alongside the two costs already moving underneath you: the July 24 tariff cliff and the August landed-cost rebuild. Margin you design in survives the discount cycle. Margin you discount away doesn't come back.

The move for your next drop

Prime Day is a volume event wearing a growth costume. The founders who win the next one aren't the ones who discount deeper — they're the ones who walked in with enough margin headroom that a promotion is a choice, not a rescue.

So before the Labor Day window opens: which of your top sellers could carry a 40% marketplace discount and still clear contribution margin — and which ones can't, no matter how many units move? If you can't answer that per SKU, that's the sheet to run this week. What would you do — cut the discount, or fix the make?

The bottom line

Prime Day didn't grow your margin — it borrowed against it, and the July settlement is the bill. A brand still reading the revenue spike as the result is measuring the wrong number.

Cost every SKU on contribution margin, know its break-even volume before you set a discount depth, and put the real work into the make — because $3 of landed cost taken out at the tech pack beats a discount weekend on every drop that follows. Do that and the next event is a lever you pull on purpose, not a hole you spend a quarter climbing out of.

FAQs