Is Black Friday worth it for my store?
Is Black Friday a dying trend?
Not by sales. Adobe measured $11.8 billion of US online spending on Black Friday 2025, up 9.1% on the year before, and $44.2 billion over the five days from Thanksgiving to Cyber Monday (Adobe). Shopify reported a record $14.6 billion in merchant sales over the Black Friday Cyber Monday weekend, up 27%, or 24% in constant currency (Shopify).
Individual stores see the lift too. In real online orders from 118 stores using TrueROAS, sales per day in Black Friday week (24 November to 1 December 2025) were a median 2.04x October. The middle half of stores ranged from 1.22x to 3.15x. Whether that extra revenue made money is a separate question, and the headlines do not answer it.
How do I know if Black Friday made me money?
Work out contribution profit: sales after discounts, minus product cost, shipping and payment fees, minus ad spend on every channel, minus what returns and refunds cost you. Then compare Black Friday week with 8 normal days, because the week in our data ran 8 days from Monday to Cyber Monday.
A made-up example shows how the same sales lift can go either way. The store has a 55% margin at full price and sells 2.04x its normal day in both Black Friday versions. We assume returned items go back on the shelf, so a return costs the margin on that order.
| Per day | Normal October day | 25% off, ads 2.5x | 15% off, ads 2x |
|---|---|---|---|
| Sales | $2,000 | $4,080 | $4,080 |
| Margin before ads | 55% | 40% | 47% |
| Margin in dollars | $1,100 | $1,632 | $1,920 |
| Lost to returns | $88 (8%) | $196 (12%) | $230 (12%) |
| Ad spend | $500 | $1,250 | $1,000 |
| Profit per day | $512 | $186 | $690 |
| Profit over 8 days | $4,096 | $1,489 | $5,517 |
Twice the revenue made 64% less profit than 8 normal days with the deep discount, and 35% more with the smaller one. In real life a deeper discount may sell more, but it has to sell a lot more. For the margin math behind this, see what discount to offer, and for spend, how much to raise your budget.
Use your own return rate. The NRF and Happy Returns survey found retailers expected 17% of holiday sales and 19.3% of online sales to come back in 2025 (NRF). Those are forecasts across all retail, so check your own December refunds.
Do the new customers make up for it later?
Less than you might hope. In the same 118 stores, a median 94% of Black Friday week orders came from first-time customers, against 85% in October. So the week does bring new buyers.
Fewer of them stay. Of the October 2025 first-time buyers, a median 10% ordered again within 90 days. Of the Black Friday week first-timers, 7% did. In 68% of 105 stores, Black Friday first-timers came back less often. Do not justify a loss-making week with repeat orders you have not seen yet. See how to judge new and returning customer ROAS.
Which stores win on Black Friday, and which lose?
My view, from working with stores, not a measured result:
| Usually worth it | Usually not |
|---|---|
| Margins of 60% or more, which can carry a real discount and still pay for ads | Margins near 30%, where 20% off needs a ROAS of 8 to break even |
| Products people buy as gifts or stock up on | Products your regular buyers simply wait to buy at the sale price |
| A large email list of people who have not bought yet | Sales that depend almost entirely on paid ads |
| Stock you would otherwise mark down later | Made-to-order products with no spare capacity |
Which numbers should I judge Black Friday by?
- Contribution profit for the week against 8 normal days, after discount, ad spend and refunds.
- Ad spend per new customer, counted on real orders.
- Refunds by early January, once most holiday returns are in.
- Repeat orders within 90 days from Black Friday first-timers, against your normal first-timers.
- Platform ROAS last. In October 2026, Meta reported a median 1.12x the sales we could match to real orders across 91 stores. A week with record revenue can still hide a channel that only took credit.
Revenue tells you Black Friday happened. These five tell you whether it paid.

