Black Friday 2026 in AI answers: what you can still change in four weeks

A new Black Friday landing page is the slowest thing you can ship this month. When a shopper asks an assistant where the best deals are, the answer is mostly built from pages that already exist: other shops, deal roundups and the media that cover the weekend. With four weeks to go, your best moves are to get onto those pages, make your own offer easy to read, and pay for the questions you can't win in time.

Last week two of my colleagues were on a call with a Nordic fashion retailer we work with. In AI answers it usually does very well: across its categories it gets recommended far more often than its biggest rival. Then we filtered to Black Friday questions in one of its markets, and it had about half the rival's visibility. Part of the reason came up a few minutes later. As the retailer put it: "due to legal reasons … we are not able to have Black Friday campaigns." Its sale runs under its own name. So when a shopper asks about Black Friday, its own site never uses the words the question is about.

You may not have that legal problem. But you probably have a smaller version of it: a campaign page that goes live the day before, a sale name AI has never seen, and product pages that don't say what's discounted.

If you sell to businesses, you can skip this one. Black Friday rarely moves B2B buyers, and a B2B prospect told us as much this month.

Why this year is different

AI assistants aren't a novelty on Black Friday anymore. Adobe expects AI-referred traffic to US retail sites to be up 95% on Black Friday 2026 compared with last year. That's after a year in which it grew many times over. It's still a small share of all traffic, and it's a US forecast. But it's the fastest-growing way shoppers arrive, and those shoppers arrive after an assistant has already given them a shortlist.

That last part is the one that matters. A search result gives the shopper ten links to compare. An assistant gives them a short answer: a few shops, a few products, often with a price and a rating beside each one. If you aren't in that answer, you don't lose a ranking position. You aren't considered at all.

What an assistant does with a Black Friday question

Take "where can I find the best Black Friday deals on winter jackets?" Before the assistant answers, it runs its own searches. These are called fan-out queries, and they're usually more specific than the shopper's question: "winter jacket Black Friday discount", "best jacket deals November 2026", a shop name plus "sale". Then it reads what comes back and writes the answer from it.

So three things decide whether you're named:

• The pages it finds. In the retailer's case, most of the pages cited for Black Friday questions were rival shops' own pages. Those are pages you can't publish on. Once you filter them out, what's left are deal roundups, gift guides and news sites. Those you can get onto.
• The words on your own pages. If the fan-out queries say "Black Friday" and "discount" and your page says "Winter Event", the assistant has to make the connection itself. It sometimes does. You don't want to depend on it.
• Your product data. In ChatGPT's shopping answers, products appear as cards with the price, the rating and the shop. That information comes from product pages and merchant listings, so a product page with no clear price or stock status gives the assistant nothing to show.

For more on fan-out queries, and how to see the ones that matter for your category, read the searches you never see.

What can still land before 27 November

AI answers don't update the moment you publish. Expect days to weeks between a change and its effect. So sort the work by how fast it lands, not by how important it sounds.

Four of those can be done by a marketing team this week. The developer work is real but small, and if your shop has a November code freeze, it goes on January's list.

Five moves for the next four weeks

1. Measure before you change anything. Write the questions your shoppers will ask, per market and per language: "best Black Friday deals on [category]", "is [your brand] having a Black Friday sale?", "where's the cheapest [product] this Black Friday?". Run them now, so you have a baseline before the weekend. Without it, you won't be able to tell what worked.

2. Make your offer readable early. Publish the sale page this week, on a URL you'll reuse every year, and use the words from the fan-out queries: "Black Friday", the category, "discount", the dates. If you can't use the name Black Friday, say clearly what the sale is and when it runs. Then let third parties make the connection for you. A page that only goes live on the 26th gives the assistant almost no time to find it.

3. Get onto the lists that already get cited. Take the pages cited in your Black Friday answers, remove the rival shops, and rank what's left by how many of your questions each one influences. Pitch the top three this week, with something an editor can use: a confirmed discount, a date, a product worth listing. This was the advice for the retailer that can't say Black Friday, and it's the fastest organic move for everyone else too.

4. Check your best sellers as a shopper would. Open the five products you most want to sell that weekend. Is the price on the page, in the code as well as the image? Is it clear that they're in stock? Are the reviews visible? These are the details a shopping answer shows. Send anything broken to a developer as a short ticket, and if it touches every page, test it before the freeze, not during it.

5. Pay for the gap you can't close in time. For your most valuable Black Friday questions where you're not named, ads inside ChatGPT are the one channel that works on a four-week timeline. Use them as a bridge while your organic work catches up, not as a replacement for it. Our guide to paid AI visibility covers what you can and can't buy.

Three mistakes we expect to see this year

• Judging the weekend by one screenshot. AI answers vary from one run to the next. Read the share of answers you appear in across the whole week, not what one person saw on Friday morning.
• Reading US growth figures as your market. The big growth figures come from the US. Your markets may be at a different stage, and they differ from each other. Measure your own before you set a budget.
• Dropping the prompts on Monday. Christmas questions start the moment Black Friday ends. Keep the same tracking running, and the work you did for Black Friday carries straight into December.

How 3RD helps with this

Most of the work above is outreach, content and a short list for a developer. 3RD shows you where to point it.

• Tracked prompts per market and language give you the baseline in step 1, with the answers and the pages each one cites.
• Fan-out queries show what the assistant actually searched for, so step 2 uses the right words.
• Outreach ranks the third-party pages by how many of your prompts they influence and shows which competitors each one already features. That's step 3.
• Product tracking (in closed beta, ChatGPT shopping answers only) shows which products and which shops the shopping answers name, with price and position, for step 4.
• Ads suggests which prompts are worth buying and lets you run OpenAI Ads campaigns from the same place. Ads tracking shows which rivals are already buying your Black Friday questions. Both are rolling out by market.

What 3RD doesn't do: it doesn't publish your sale page, pitch the editors or change your product templates. Your team does that. And it doesn't claim that one task caused a sale. It shows the trend next to the work you shipped.

About the data

• Adobe's 2026 forecast: AI traffic to US retail sites expected up 130% year over year for the holiday season (1 November to 31 December), and up 95% on Black Friday. Adobe counts AI traffic as shoppers who click a link from an AI chat or browser service to a retail site. Source: Adobe's 2026 holiday forecast, reported by Digital Commerce 360. Adobe sells analytics and commerce software to retailers.
• Last year: Adobe reported that AI traffic to US retail sites was up 805% on Black Friday 2025, on a day with a record $11.8 billion in US online spending (Reuters, via U.S. News).
• The retailer example comes from a customer call in October 2026. The company isn't named, and the quote is verbatim.

FAQ

What if we can't use the words "Black Friday"? Describe the sale clearly in your own words, then rely on third-party pages that do use the name: deal roundups, gift guides and media coverage. AI can connect your sale to the question when other sources make the link for it.

Next step

This week, write ten Black Friday questions for your most important market and run them. Note who gets named, and which pages get cited. That tells you whether your four weeks belong to outreach, your own pages or ads.

If you'd like to see those answers per market, with the cited pages and the rivals already buying ads, start for free and set up your Black Friday questions in 3RD.

Magnus Windeløv

Head of Marketing

Magnus leads marketing at 3RD. He has spent his career in demand generation, growth and SEO for B2B software companies and consumer brands.

View profileLinkedIn