Black Friday without the discount

This autumn I designed the Black Friday offer for an apparel brand in golf and corporate wear. Earlier in the year it had walked away from the 30-35% sitewide promotions it used to run, and by August orders were growing again at full price, up 8.3% (count, YoY) at a flat average order value.

The offer kept every product at full price through Cyber Monday and added a gift card for the next visit. Spend $150, get a $20 card. Spend $300, get $35. I sized it from the store's own order data and built the plan around it, and the agency team handled the channels. The offer below is the design as I handed it over, not a report on how it ran.

What two years of 30% off bought

The brand ran the same Black Friday in 2024 and 2025: 30% off sitewide, with early access for its most loyal customers. In 2025, 95% of orders in the window (count) used the discount, and it came to 29.1% of the list value of everything sold ($).

Orders in the window fell 29.5% anyway (count, 2025 against 2024), and November and December revenue fell 24% ($, YoY). The discount was on nearly every order and didn't hold the volume.

The brand had spent this year getting customers used to full price again, and a 30% Black Friday would have taught them to wait for it. So the question became what to give them in place of a discount.

Sizing it on full-price baskets

My first version put the rungs at $200 and $300. Last Black Friday's orders seemed to support that: the median order was $150 at list price, and 33% of orders (count) reached $200. Those baskets were built at 30% off, though, and this year there wouldn't be one.

So I sized it on August, the only full-price month on record. At full price the median order was $115 and the average item $83, at 2.1 items per order. Of those orders, 41% (count) came in under $100, only 20% reached $200, and 9% reached $300.

A $200 first rung would have given nothing to four buyers in five. The largest single band, $100 to $149, held 26% of orders (count), and any order in it clears $150 with one more average item. It's twice the size of the band just below $200, which held 13% (count).

So the first rung moved to $150, about one item above the median, where 34% of full-price orders (count) already landed. The $300 rung stayed where it was, around where the top tenth of orders begin.

What the gift does to margin

A percentage off costs its full rate on nearly every order. A gift card costs nothing until someone spends it, and only the orders that reach a rung earn one. I planned on 40% of promotional cards (count) being redeemed, so a $20 card costs about $8 for every one issued.

I modelled the options on the August orders, assuming a 55% gross margin (the brand hadn't shared its own figure, and the ranking held from 45% to 65%) and that three in ten orders sitting one item below a rung would add that item. The $150 and $200 versions kept almost the same margin per order. The $150 version gave a card to 34% of orders (count) against 20%, and its basket lift per order came out about three quarters larger ($).

Last year's discount, even on its bigger baskets, kept about half the margin per order that the tiers did, and the tiers' expected cost per order was roughly a twelfth of the discount's ($, modelled). They matched last year's total margin at just over half of last year's orders (count). In a floor case where orders fell another 29% (count, YoY), modelled revenue came within 4% of last year's ($) and margin rose 38% ($).

Of the other structures I costed, a flat $20 card on every order is the easiest to explain, but it pays about $8 on a $40 pair of socks and asks nothing of the basket. Buy one, get the second half price is a 25% sitewide discount under another name.

The card is a fixed amount, so nothing on the site is ever marked down. $20 on a $150 order is 13% of the order value, and $35 on a $300 order is 12%.

What has to be in place

The rules sit on the second order. The $20 card needs an order of $100 or more within 45 days, the $35 card one of $200 or more, so a card is never worth more than 20% of the order it's spent on. Each qualifying order earns one code, sent when it ships and cancelled if it's refunded. The 45-day window runs into January, the brand's quietest month, so the return visit can land there at full price.

A progress line in the cart that counts up to $150 and $300 is where the second item gets added. Without it I'd take the basket lift out of the model.

On this brand's store platform, gift certificates couldn't be issued by an automatic promotion. So the codes come from a post-purchase flow in the email platform, with a branch at each threshold.

Spend-and-get offers are common at value retailers, so the line is "a $20 gift card for your next visit", never "cash back".

Do this Monday

Your Black Friday orders are a week old. Before you plan the next sale, size any gift tiers from orders placed at full price, because a discount inflates the basket you'd be measuring.

1. Export a full-price month. Pick a recent month with no sitewide code and take each order's value at list price.
2. Group the orders by value in bands of about fifty dollars, and note the median order and the average item price.
3. Set the first rung about one average item above the median, and check that the band just below it is one of your largest.
4. Set the second rung around where your top tenth of orders begin.
5. Cost each rung: card value, times expected redemption, times the share of orders (count) that land on it. Put the total next to what this year's discount cost as a share of list price.
6. Write the redemption rules before the creative: a minimum on the next order, a window, and one code per order.

---

In one line: size a Black Friday gift from what customers spend at full price, with the first rung about one item above the median order.