Every Black Friday playbook is built on the same assumption: shoppers see your offer before they see anyone else’s. That’s why retailers invest in doorbusters, homepage takeovers and countdown timers. They grab attention first and let urgency do the rest.
That assumption is starting to break. This year, more shoppers will ask an AI assistant to find them the best deal before they visit a single store or website. The assistant may read your banner, but a countdown timer doesn’t rush it and a doorbuster doesn’t tempt it. It just compares prices.
That sounds like a small change, but it affects how much your Black Friday discounts are really worth. To see why, it helps to look at what a discount actually does for a retailer.
What a Black Friday discount is really paying for
When a retailer runs a big Black Friday discount, it expects that discount to do two separate things.
First, it should win on price. A shopper compares the product across a few stores, sees yours is cheaper, and buys from you. This is the obvious job of a discount.
Second, it should bring people in. A big discount on the homepage, in an email subject line or on a “doorbuster” sign gets attention. Shoppers who see it come to your site or store, look around, and often buy other things too, even if the discounted item itself wasn’t the cheapest on the market (this effect is also referred to as “halo effect”). Here the discount is working less like a price and more like an ad.
Most retailers think about their Black Friday discounts mainly in terms of the first job. But the research suggests the first job often doesn’t pay off on its own. BCG found that 20% to 50% of retail promotions don’t noticeably increase sales of the discounted product, and another 20% to 30% increase sales but not by enough to cover the cost of the discount. McKinsey found that even the best grocers can expect 10% to 15% of their promotions to hurt both sales and margins, once you count shoppers who would have bought anyway, or who simply switched from another product you sell.
Discounts are getting deeper anyway. In the 2025 holiday season, Adobe tracked average online discounts of nearly 31% in electronics, about 30% in toys and about 25% in apparel, deeper than the year before in most categories.
So why do retailers keep running so many discounts that don’t pay back on price? Because of the second job. Even when a discount loses money on the product itself, it still brings shoppers through the door, and on the busiest weekend of the year that traffic feels worth paying for. Attention has been quietly propping up a lot of Black Friday discounting.
That’s the part AI changes. When a shopper asks an AI assistant to find the best deal on a TV, the assistant works mostly from product data: prices, availability and delivery details, increasingly pulled through structured feeds and agentic commerce protocols. Even when it reads your homepage banner, it treats it as one more data point, not as a reason to pay attention. It extracts the price, compares it and gives the shopper a shortlist. Your discount can no longer bring that shopper in just by being loud. It only helps if it actually wins the price comparison. The second job fades, and you’re left with the first one, the one that was already the weaker of the two.
How many shoppers are actually using AI to shop?
Not everyone, but the number is growing fast.
- It’s already shaping real purchases. Salesforce found that AI and AI agents influenced 20% of all global orders during Cyber Week 2025, worth $67 billion in sales.
- It’s sending more shoppers to retailers. Adobe saw traffic from AI tools to US retail sites rise nearly 700% during the 2025 holidays. That growth started from a small base, but it’s moving quickly.
- Shoppers trust it. BCG reports that shopping-related use of AI tools grew 35% in nine months of 2025, and more than 60% of consumers say they highly trust the results.
- More shoppers will start there this year. In Bain’s survey, 24% of US shoppers plan to begin their holiday shopping on AI platforms like ChatGPT, Gemini and Claude, up from 17% last year.
BCG sums up what this does to the sale season well: deal hunting is becoming “more professionalized and data-driven.” In plain terms, more shoppers are letting a machine do the comparison for them.
Most shoppers will still find you the usual way this year. But the ones using AI are likely to be exactly the shoppers retailers compete hardest for on Black Friday: people who know what they want and are looking for the best price.
So when is a discount still worth it?
If a discount can’t rely on bringing people in the way it used to, it has to earn its place some other way. There are really only three good reasons to discount a product on Black Friday:
- It wins a price comparison you’d otherwise lose. Shoppers, or their AI assistants, are comparing this product across stores, and your price is what tips the decision your way.
- It clears stock you need to move. You have too much of the product, or it’s near the end of its season, and selling it now at a discount is better than marking it down further later.
- It shapes how affordable you seem overall. Some products are the ones shoppers notice and remember, and their prices shape what people think of your whole store. A good price on those makes everything else you sell feel like better value.
Most Black Friday discount plans include plenty of markdowns that don’t fit any of these three reasons. A product gets discounted because it was discounted last year, or because the whole category is on sale, or because nobody stopped to ask whether it needed to be. Often it’s a product shoppers weren’t comparing, that you weren’t overstocked on, and that would have sold at full price anyway.
This happens more than most retailers realize. BCG found that one of the main reasons promotions fail is that retailers often just repeat last year’s promotions instead of planning them around a clear goal.
In the past, these discounts were easy to justify, because every discount brought some attention with it. Even if the price cut itself didn’t pay off, the banner still pulled shoppers in. When AI does the price checking, much of that attention goes away. A discount that doesn’t win a comparison, clear stock or shape your price image is no longer buying you anything. It’s just a margin you’ve given away.
So the useful question for this year isn’t “how deep should we go?” It’s “which of our discounts are actually doing one of these three jobs, and which are just running because they always have?”
Sorting your catalog: which products get compared?
A simple way to answer that is to sort your products into three groups, based on how shoppers judge them.
Group 1: The products everyone compares. These are the items shoppers ask AI assistants about and use to decide whether you’re competitive, like the TV, the laptop or the popular toy. McKinsey calls these key value items: the products “whose prices consumers tend to notice and remember.” These need to win the comparison and keep winning it as competitors change prices through the weekend. This is where your discount budget should go first.
Group 2: The products bought alongside them. These are the cables, cases, accessories and add-ons that end up in the same basket. Shoppers rarely check their prices one by one, so discounting them as deeply as your headline products usually gives away margin for nothing. Keep an eye on this group, though: a shopper might not bother checking the price of an HDMI cable, but an AI assistant can.
Group 3: The products only you sell. These are exclusive lines, private-label products, or items that are hard to find a substitute for. Since shoppers can’t easily compare them, you can protect their margins. Just be honest about which products really belong here. If an AI search turns up a close alternative, the product isn’t as unique as you think.
One more thing to keep in mind: these groups change. A shopper can only remember a few prices, but an AI assistant can check far more, in seconds. So the list of products that shapes your price image is getting longer, and it shifts with what people are asking about. A product nobody compared in July can become one of the most compared items by late November. That’s why the groups need to be reviewed before every big event, not set once a year.
Don’t forget about stock
A discount only works if you have the stock to back it up. AI assistants look at availability as well as price (more on how they choose), so stock problems show up quickly. Two mistakes come up every year.
Discounting products that would have sold out anyway. If a product is going to sell through at full price, cutting the price just gives margin to shoppers who would have paid more. It can also leave you out of stock for the rest of the holiday season.
Advertising a great price on something you can’t deliver. Backorders and cancellations don’t just cost you the sale. They make you a less likely pick the next time a shopper, or their assistant, compares.
The products where a deep discount makes the most sense are the ones you have too much of. That’s why competitor-aware pricing works best when it’s also stock-aware: what your competitors charge matters, but so does how many units you have left to sell.
Change prices during the event, but set your limits first
Prices change constantly over Black Friday weekend. When a competitor drops its price on a popular product, AI assistants can pick that up quickly, so a pricing plan you locked in weeks ago won’t hold up.
But reacting to every competitor move isn’t the answer either. If everyone keeps matching everyone, prices across the whole category fall, and nobody makes money.
The solution is to make two decisions before the weekend starts:
- Which competitor moves will you actually respond to? Some competitors affect your sales in some categories. Most of their price changes don’t. Decide in advance where you’ll react and where you’ll hold your price.
- What are your limits? Set the lowest and highest prices you’ll accept, and the minimum margin you need, for each group of products. Pricing guardrails like these let your team move quickly during the event without ending up with a margin problem afterward.
Look beyond the weekend
Black Friday is the biggest moment of the season, but it isn’t the whole season. Salesforce called Cyber Week the most important purchasing window of the year, and shoppers keep comparing prices long after it ends.
Bain’s data shows that six of the top 10 sales days in 2025 fell in the final two weeks before Christmas. Christmas falls on a Friday this year, so that last week could be especially busy. Any margin you spend on discounts that aren’t doing a job in November is margin you won’t have for the shoppers still comparing in December.
Before Black Friday: a three-question test
Before you finalize your Black Friday discounts, run each one through three simple questions:
- Will anyone compare this price? If shoppers and AI assistants aren’t checking it, a deep discount won’t buy you much.
- Will it win the comparison? A discount that still loses to a competitor costs you margin and doesn’t win the sale.
- Can we deliver it? A winning price on a product you’re about to run out of costs more than it earns.
If the answer to any of these is no, and the product doesn’t need clearing, the discount probably isn’t earning its cost. Cut it, and spend that margin somewhere it will actually do something.
Key takeaways
- A Black Friday discount does two jobs: it wins on price, and it brings shoppers in. AI assistants are taking away much of the second job.
- Research from BCG and McKinsey shows that many promotions didn’t pay back on price even before AI. The attention they brought was covering for them.
- AI influenced 20% of global Cyber Week orders in 2025, according to Salesforce, and more shoppers are letting AI do their deal hunting.
- A discount is only worth it if it wins a price comparison, clears stock or shapes how affordable you seem. If it does none of these, it’s costing you margin for nothing.
- Sort your products by how they get compared, factor in stock, set your pricing limits before the event, and save some margin for December.
FAQ
How is AI changing Black Friday shopping?
More shoppers are asking AI assistants to compare prices for them before they visit a store or website. Salesforce found that AI and AI agents influenced 20% of global orders during Cyber Week 2025, and BCG reports that shopping-related use of AI tools grew 35% in nine months of 2025.
How can retailers tell if a Black Friday discount is worth running?
Ask three questions: will shoppers or AI assistants compare this price, will the discount win that comparison, and do you have the stock to deliver it? If the answer to any of them is no, and the product doesn’t need clearing, the discount is likely giving away margin without bringing in extra sales.
Do Black Friday promotions actually pay off?
Often not. BCG found that 20% to 50% of retail promotions don’t noticeably increase sales, and McKinsey estimates that even the best grocers see 10% to 15% of their promotions hurt both sales and margins.
How should retailers set Black Friday prices when shoppers use AI assistants?
Put your best prices on the products shoppers actually compare, keep them competitive as rivals change prices, check your stock before discounting, and set pricing limits in advance so you can react quickly without hurting margins.
Does Black Friday still matter if the holiday season is getting longer?
Yes. It’s still the peak of the season. But the final two weeks before Christmas matter a lot too, so plan your margin for the whole period, not just the weekend.
How Quicklizard helps retailers price for the comparison era
Quicklizard is an AI-powered dynamic pricing platform built for omnichannel retailers. It helps pricing teams tell the discounts that do a job from the ones that only cost margin:
- Pricing by product role. Quicklizard sets each product’s pricing strategy according to its job, whether that’s a key value item that shapes price perception, a sales driver or a profit generator. Discounts go where comparison happens, and margin is protected where it doesn’t.
- Competitive Pricing Intelligence. Quicklizard monitors competitor prices and analyzes how they affect your demand, so teams can respond to the competitor moves that matter during peak events and ignore the rest.
- Promo Effectiveness. Quicklizard shows which discounts are adding sales and which are only giving away margin, so teams can keep the promotions that earn their cost and cut the ones that don’t.
- Glass Box AI. Every pricing recommendation comes with the reasoning behind it, so teams can review and trust decisions made at Black Friday speed.