BFCM 2026: the discount math that decides whether November makes you money

Every November, a store owner faces the same question: what do I put on the site?

The reflex answer is a percentage. Twenty percent off everything, sitewide, one code, easy to explain in an email subject line. It is also the most expensive way to run Black Friday, and most stores never do the arithmetic that shows why.

This is that arithmetic. It is worked with round numbers so you can substitute your own. And because Black Friday 2026 falls on Friday 27 November — with Cyber Monday on Monday 30 November — there is a calendar at the end that counts backwards from those dates.

What the season actually looks like this year

Two sets of numbers set the context.

Shopify's BFCM 2025 recap: merchants processed $14.6 billion globally over the weekend, up 27% on 2024. More than 81 million customers bought from Shopify-powered stores, 94,900+ merchants had their best sales day ever, and at the peak — 12:01pm Eastern on Black Friday — the platform was processing $5.1 million per minute. The average cart reached $114.70, up from $110.71 the year before.

Adobe's 2026 forecast, published 28 September: US online holiday spending of $275.1 billion, up 6.7%. Cyber Week accounts for $47.5 billion of that, up 7.4%, with Black Friday at $12.9 billion (+9.2%) and Cyber Monday at $15.1 billion (+6.2%).

Two things follow from that. First, carts are already big in this window — the average Shopify cart is over a hundred dollars, which means the order value you are defending is not trivial. Second, Black Friday is growing faster than Cyber Monday. Demand is moving earlier. If your offer only exists for 24 hours on the Saturday, you are running the wrong week.

The trap: a discount rate is not a cost

Here is the sentence most stores never write down. The headline percentage is what the shopper sees. The blended percentage is what you pay.

A sitewide code applies to every order — including the orders that were going to happen anyway, at full price, from customers who had already decided to buy. You are handing a discount to people who did not ask for one.

Tiered offers work the other way. The headline is bigger — "buy 5, save 20%" sounds more generous than "20% off" — but it only pays out when the shopper adds more to the basket. The discount is conditional on behaviour that raises the order value. So the number that lands on your P&L is much smaller than the number on the banner.

A worked example

Take a store doing $40 average order value at a 60% gross margin, with 1,000 orders over the promotion. Cost of goods is 40% of list price.

The assumptions behind the tiered column: 65% of orders take no offer; the other 35% take a tier averaging 14% off and buy 2.8 units instead of one. Notice what the blended discount does — a headline of "up to 20%" only costs 8.4% across the whole promotion, because most orders never claim it.

Now the numbers most people skip past. The sitewide column cuts gross profit by a third ($24,000 to $16,000) to produce less revenue than doing nothing in the tiered column. That is the whole argument in one line: a sitewide percentage buys you a smaller basket at a lower margin.

These are illustrative inputs, not a benchmark. Run your own:

Blended discount = total discount given ÷ list value of everything sold
Cost of the promotion = revenue after discounts − cost of goods

The second formula matters more than the first. A promotion that grows revenue while shrinking contribution margin is not working; it is buying sales at a loss.

Which tier structure fits which catalogue

The shape of the offer depends on what the shopper is buying. Consumables reward depth; considered purchases reward bundling.

Two practical constraints. Keep it to three tiers at most — a fourth tier is decoration and shoppers stop reading. And put the top tier roughly 1.5 to 2 times your current average order value, close enough to reach, far enough to be worth reaching.

The calendar, counting back from 27 November

The dates that actually matter are not the sale dates. They are the testing deadlines.

The reason for the 20 November freeze is simple: new apps and new code installed during the sale week break in front of your largest audience of the year, and you find out at 12:01pm on Black Friday when nobody is watching the support inbox.

Where the offer has to appear

A tier that only exists at checkout is invisible. The shopper never discovers it, so the behaviour you are paying for never happens.

Three placements, in order of importance:

1. On the product page, next to the variant picker or quantity selector. This is where the decision to buy two instead of one is made.
2. In the cart and cart drawer, as a line that shows what the next tier saves.
3. In the announcement bar or a collection banner, at a lower priority — useful for shoppers who arrive cold.

If you use Quantity Breaks, the tiers render as an app block on Online Store 2.0 themes, so they install from the theme editor without touching code. Discounts are applied through Shopify Functions, which matters this year: Shopify Scripts was sunset on 30 June 2026, so any store still relying on a Script-based discount is relying on something that no longer runs. If your current discount app has not moved to Functions, that is not a BFCM optimisation — it is a fault you will discover during the sale.

How to measure it honestly

Average order value is the headline metric and the least useful one on its own. A discount can lift AOV while destroying the difference between it and profit.

Track these four:

• Contribution margin per order — revenue minus discounts minus cost of goods. This is the number the promotion either grows or shrinks.
• Units per order — the tier's actual job. If this does not move, the tiers are not being seen.
• Blended discount rate — total discount ÷ list value. If this sits near your headline percentage, nobody is taking the tiers; they are just taking the discount.
• Take-rate per tier — which tier shoppers actually claim. If tier three is never reached, the top step is set too high.

The right comparison is not "BFCM this year versus BFCM last year". It is "this promotion versus the same traffic with no promotion", which is why testing before 6 November matters — afterwards you will never have a clean baseline again.

What not to do

• Do not stack a sitewide code on top of tiers. The blended discount compounds and the margin disappears.
• Do not extend the end date. A sale that gets extended teaches your list that the deadline is not real, and the next one converts worse.
• Do not discount your hero product to the deepest tier. Protect the margin on the item that already sells; put the depth on the attach.
• Do not launch a new theme or a new app in sale week. Freeze from 20 November.
• Do not judge the result on revenue alone. Revenue is the easiest number to grow and the easiest to grow unprofitably.

Getting it running

If you want the mechanics of tier design in more depth — where on the page each tier should sit, how to name them, how to measure the lift without fooling yourself — that is covered in How to increase AOV with quantity breaks.

If you want the adjacent levers, they pair cleanly with tiers and cost nothing on the margin side: a free shipping progress bar that counts down to the threshold, and a free gift threshold instead of another percentage off.

The arithmetic in this post is the part worth doing before anything is installed. Write down your margin floor, work out the blended discount each structure would produce at your current conversion rate, and pick the shape where the promotion still makes money. That is a spreadsheet exercise that takes an hour — and it is the difference between a November that grows the business and one that only grows the revenue line.

Have a specific catalogue and no clear answer on the tier depths? Send us the numbers and we will tell you what the blended discount would look like before you commit.