What AI tells shoppers in six countries, from 360 questions on four AI surfaces
Opttab Research. Data collected 4 November 2026.
In early November, we asked AI 360 shopper questions about Cyber Monday: 60 in each of six countries (United States, Canada, United Kingdom, Germany, France and Türkiye), each in the local language. Every question went to four AI surfaces: ChatGPT, Gemini, Google AI Overview and Google AI Mode. That gave 1,440 search-and-surface pairs and 1,409 answers with text. Google showed an AI Overview on 329 of the 360 searches. We read every answer to see what a shopper actually gets.
This is our second holiday study. The first, on Black Friday, was collected on 31 October 2026 and covered 600 prompts on five surfaces. The two studies use different prompts, surfaces and dates, so any comparison between them is directional, not like-for-like.
The short version: AI tells shoppers not to wait for Cyber Monday. It gives specific price evidence on Cyber Monday deals less often than in our Black Friday study. Europe hears “Cyber Week”, Türkiye hears its own November. And AI still won’t buy things for you.
Key findings
Buy on Friday, not Monday. Asked whether to buy on Black Friday or wait for Cyber Monday, 38.5% of answers say Black Friday. Only 7.3% say wait for Cyber Monday. 31.8% say it depends.
Weaker deal checks. Asked whether a Cyber Monday deal is real, 38.7% of answers give specific price evidence (Black Friday study: 74.2%). 56.5% fall back on generic tips.
Cyber Week in Europe, Türkiye’s own calendar, little Christmas. 37.6% of UK, German and French answers frame the sales as Cyber Week (Germany 48.9%). In Türkiye, Cyber Week appears in 1.7% of answers, while 44.6% mention Efsane Cuma, 11.11 or Kasım sales. Christmas delivery comes up in only 3.4% of all answers.
Canada still lags. Only 38.3% of retailers named to Canadian shoppers are Canadian (Black Friday study: 37.5%). 10.1% of priced Canadian answers use a foreign currency, and only 35.9% of citations come from Canadian sites.
Community in North America, comparison sites in Germany and Türkiye. Reddit, Facebook and YouTube are the three most-cited sites in the US and Canada. idealo.de is cited in 11.3% of German citing answers, akakce.com in 18.3% of Turkish ones.
AI says it can’t buy. Asked to buy something, 82.1% of answers say they can’t (Black Friday study: 64.4%). A strict keyword check finds 1 of 162 answers offering cart or checkout.
What it means: the pages, feeds, event names and policies AI can read will decide whether it sends shoppers to you on 27 November, on 30 November, or not at all.
The numbers at a glance
Number | What it measures |
|---|---|
1,409 | AI answers with text, from 360 searches × 4 surfaces (1,440 slots; AI Overview shown on 329 of 360) |
38.5% | of timing answers say buy on Black Friday |
7.3% | say wait for Cyber Monday |
38.7% | of “is this Cyber Monday deal real?” answers give specific price evidence (Black Friday study: 74.2%) |
44.6% | of Turkish answers mention Efsane Cuma, 11.11 or Kasım sales |
3.4% | of answers mention Christmas delivery |
82.1% | of “buy it for me” answers say they can’t buy |
28.7% | chance two AI answers to the same question share the same top pick |
Finding 1: AI says buy on Black Friday
In every market we asked: should I buy on Black Friday, or wait for Cyber Monday? Across 192 answers:
Advice | Share |
|---|---|
Buy on Black Friday | 38.5% (cluster 95% CI 19.3 to 58.9%) |
It depends | 31.8% |
No timing advice | 17.2% |
Wait for Cyber Monday | 7.3% |
Wait until later | 5.2% |
Europe leans hardest toward Friday. In Germany, 59.4% of timing answers say Black Friday, and none of 32 say Cyber Monday. The UK is at 50.0%. American answers hedge most: 53.1% say it depends, 28.1% say Black Friday and 3.1% say Cyber Monday. Canada gives Cyber Monday its best showing, at 15.6%.
Dates are right when given. 19.6% of answers state a Cyber Monday date, and 273 of 276 are correct (30 November). Black Friday dates appear in 7.3% of answers, and 102 of 103 are correct (27 November). Timing answers state the Cyber Monday date 38.5% of the time, and 75.0% in Türkiye. Outdated answers are rare: 2.6% present past-year deals as current or give a wrong 2026 date (Black Friday study: 1.0%), most often in the US (5.6%).
Why it matters: if AI frames Friday as the moment to buy, a Cyber Monday offer needs a reason to wait that AI can read and repeat. Without that reason, shoppers AI sends on Monday may already have bought on Friday.
Finding 2: Fewer specific deal checks
In our Black Friday study, AI often acted like a skeptical shopper. Here it does that less often. Across 191 answers to price and deal-validity prompts, 38.7% gave specific evidence such as a price history, a comparison with other stores or a named price tracker (cluster 95% CI 21.1 to 59.9%). 56.5% gave generic tips (“compare prices”, “check reviews”), and 4.7% ignored the question. In the Black Friday study, specific evidence reached 74.2%.
Country | Specific deal evidence |
|---|---|
UK | 56.2% |
FR | 54.8% |
US | 46.9% |
DE | 34.4% |
TR | 25.0% |
CA | 15.6% |
Canada stands out. 70.3% of Canadian deal-check answers gave specific evidence in the Black Friday study, against 15.6% here, and 84.4% now give only generic tips. In Türkiye, 21.9% of deal-check answers don’t address the question at all.
Fewer concrete prices. 54.2% of answers state a concrete price (Black Friday study: 75.9%), from 64.9% in France to 45.4% in Türkiye. Across all answers, 33.5% carry any specific price evidence (Türkiye 22.9%). Caveats that prices aren’t live yet appear in 8.9% of answers (Black Friday study: 5.9%), highest in Canada at 13.7% and lowest in the UK at 4.3%.
Why it matters: with fewer specific checks, an honest Cyber Monday discount gets less automatic validation, and a weak one gets less scrutiny. Brands that publish their own price history give AI evidence it isn’t finding elsewhere.
Finding 3: Cyber Week in Europe, Türkiye’s own November, and little Christmas
AI adapts the event to the market. 37.6% of UK, German and French answers frame the sales as a Cyber Week rather than a single Monday: Germany 48.9%, France 35.1%, UK 28.9%. Even in the US, 31.5% of answers mention Cyber Week.
Türkiye runs on its own calendar. Cyber Week appears in just 1.7% of Turkish answers (4 of 240), while 44.6% mention Efsane Cuma, 11.11 or Kasım sales (107 of 240). Turkish answers are also the most likely to explain how Cyber Monday relates to the local market: 19.2%, against 2.6% in Canada, 3.4% in the UK, 3.3% in France and 5.2% in Germany. Canada shows the opposite risk: 7.7% of Canadian answers assume a US context (18 of 233). Few answers anywhere describe Cyber Monday as an online-focused event (7.2%).
Christmas barely comes up. For many shoppers, Cyber Monday is the last big sale before gift deadlines, yet only 3.4% of all answers mention Christmas delivery. Even on delivery and returns prompts it reaches only 25.1%:
Country | Christmas delivery mentioned (delivery and returns answers) |
|---|---|
CA | 46.9% |
DE | 32.3% |
US | 25.0% |
UK | 21.9% |
FR | 12.5% |
TR | 12.5% |
Policy answers are concrete. 69.1% of delivery and returns answers give a specific returns window or cut-off (Black Friday study: 72.1%), led by the US at 87.5% and Türkiye at 84.4%, with Canada lowest at 53.1%. Google’s AI Overview appears on 97.9% of these searches.
Why it matters: the name of your event decides whether AI connects your offer to the shopper’s question. Where AI rarely answers something, such as Christmas arrival, your own pages can fill the gap.
Finding 4: Local in language, not always in stores
Every one of the 1,409 answers came back in the shopper’s language, and 97.9% of retailer mentions (2,831 mentions of retailers not named in the prompt) point to a retailer with a storefront serving the shopper’s country. Only 66.4% are home-grown, meaning headquartered in that country (Black Friday study: 66.1%).
Country | Home-grown retailer share |
|---|---|
US | 88.8% |
TR | 76.7% |
UK | 68.1% |
FR | 64.6% |
DE | 61.8% |
CA | 38.3% |
Canada remains the outlier. 10.3% of Canadian answers name a retailer with no Canadian storefront. 10.1% of priced Canadian answers use a foreign currency (2.4% overall). Only 35.9% of citations in Canadian answers come from Canadian sites (Black Friday study: 42.8%), and 33.4% come from other countries’ sites.
Who gets named:
Country | Most-named retailers (share of answers) |
|---|---|
US | Walmart 24.1%, Amazon 22.0%, Best Buy 16.8%, Target 16.4% |
Canada | Best Buy 22.7%, Amazon 20.6%, Walmart 17.2% |
UK | Amazon 22.4%, Argos 17.7% |
Germany | Amazon 18.0%, MediaMarkt 15.5% |
France | Amazon 21.8%, Fnac 15.1% |
Türkiye | Trendyol 29.6%, Hepsiburada 26.7%, Amazon 7.1% |
Türkiye is the most concentrated market: the top three retailers take 37.7% of mentions, against 21.4% in France.
Why it matters: in Canada, domestic retailers still compete with US content that AI treats as relevant. Elsewhere, being local is expected; getting named is the contest.
Finding 5: Community voices in North America, price comparison in Germany and Türkiye
92.2% of answers cite at least one source, with 3.98 cited domains per answer on average.
Source type | Share of citations |
|---|---|
Publishers and review sites | 29.1% (Black Friday study: 27.5%) |
Retailers | 24.2% (Black Friday study: 27.1%) |
Brands | 10.9% |
Video and social | 6.3% |
Price comparison | 5.8% |
Forums | 4.8% |
Marketplaces | 3.9% |
Consumer organizations | 3.7% |
Other | 11.3% |
The mix is national. Publishers carry 34.1% of US citations but 21.0% in Türkiye. Video and social reach 9.0% in Canada and 7.5% in the US. Price comparison sites reach 7.6% in Germany and 9.7% in Türkiye, where marketplaces add 10.0%. Consumer organizations reach 6.5% in the UK and 6.0% in France.
Country | Most-cited domains (share of citing answers) |
|---|---|
US | reddit.com 16.3%, facebook.com 13.4%, youtube.com 12.4%, bestbuy.com 9.9%, walmart.com 9.4% |
CA | reddit.com 19.6%, facebook.com 14.6%, youtube.com 11.4%, bestbuy.ca 10.0%, google.com 8.7% |
UK | google.com 12.3%, which.co.uk 11.8%, reddit.com 11.3%, moneysavingexpert.com 10.8%, facebook.com 9.9% |
DE | mediamarkt.de 14.0%, youtube.com 12.7%, idealo.de 11.3%, reddit.com 8.6%, businessinsider.de 6.8% |
FR | reddit.com 11.8%, facebook.com 10.9%, lesnumeriques.com 10.4%, ouest-france.fr 9.0%, idealo.fr 8.1% |
TR | akakce.com 18.3%, trendyol.com 18.3%, instagram.com 13.8%, hepsiburada.com 11.6%, sikayetvar.com 8.0% |
Why it matters: in North America, AI builds its Cyber Monday view from what shoppers say to each other on Reddit, Facebook and YouTube. In Germany and Türkiye, it reads price comparison listings. In the UK, it reads Which? and MoneySavingExpert. Your own site is one voice among several.
Finding 6: AI hands off the purchase
Across 162 “buy it for me” answers:
Level | What the answer did | Cyber Monday study | Black Friday study |
|---|---|---|---|
0 | No buying help | 11.7% | 8.3% |
1 | Points to a store or page | 38.3% | 46.8% |
2 | Shoppable listings or product cards | 35.8% | 33.9% |
3 | Offers to act (cart, checkout, price tracking or alerts) | 14.2% | 10.9% |
82.1% say outright that they can’t complete the purchase (Black Friday study: 64.4%). Declines are near-universal in Türkiye (96.4%) and Germany (96.2%), and lowest in the UK (66.7%). In Canada, they reach 84.6% (Black Friday study: 56.9%).
A strict keyword check for cart or checkout offers finds 1 of 162 answers (0.6%). Across all answers, 3.7% offer any action such as buying, tracking a price or setting an alert, and 6.6% show native product cards (US 13.4%, Türkiye 1.7%).
Why it matters: for a logged-out shopper, AI hands the purchase off and says so plainly. The merchant with a connected, machine-readable path to checkout is best placed to receive that hand-off.
Also worth knowing
AI Overview is almost always there. Google showed an AI Overview on 91.4% of searches (Black Friday study: 83.0%), including all 48 timing searches. It steps back most on gift finding (76.2%) and deal discovery (81.2%). By country it ranges from 86.7% in the US and UK to 100.0% in Türkiye. Searches that say Cyber Monday showed one 92.5% of the time, Cyber Week searches 88.9%.
There is no single “AI rank”. Two answers to the same question share the same top pick only 28.7% of the time. All top picks agree on just 17.7% of prompts, and 10.3% of answer pairs share no brands at all.
Budget barely changes specificity. Answers name 1.83 / 1.83 / 2.03 products on average for low / mid / high budgets, and state a price 55.0% / 54.5% / 53.2% of the time.
Lists are short. Answers name 1.9 products on average, and 34.3% name three or more.
Policy detail is rare outside policy questions. Across all answers, 10.3% give a concrete returns window or cut-off.
Six countries at a glance
Country | Home-grown retailers | Home-market citations | Local-currency prices | Specific deal evidence | Advice: buy on Black Friday | Declines to buy |
|---|---|---|---|---|---|---|
United States | 88.8% | 61.5% | 98.5% | 46.9% | 28.1% | 74.1% |
Canada | 38.3% | 35.9% | 89.9% | 15.6% | 31.2% | 84.6% |
United Kingdom | 68.1% | 64.5% | 100.0% | 56.2% | 50.0% | 66.7% |
Germany | 61.8% | 73.9% | 99.1% | 34.4% | 59.4% | 96.2% |
France | 64.6% | 66.0% | 100.0% | 54.8% | 28.1% | 75.0% |
Türkiye | 76.7% | 72.8% | 98.2% | 25.0% | 34.4% | 96.4% |
Country figures for deal checks, timing and declines rest on 26 to 32 answers each, and many confidence intervals overlap. Read them as indicative.
United States: the community market. 88.8% of named retailers are American. Reddit (16.3%), Facebook (13.4%) and YouTube (12.4%) are the most-cited sites, and publishers carry 34.1% of citations. Timing answers hedge: 53.1% say it depends. Local play: be present in the threads, groups and videos AI reads, and state clearly which offers run through Cyber Monday.
Canada: the localization gap. Only 38.3% of named retailers are Canadian, only 35.9% of citations come from Canadian sites, and 7.7% of answers assume a US context. Canada is where AI most often recommends Cyber Monday (15.6%) and most often mentions Christmas delivery on policy questions (46.9%). Local play: make “.ca”, CAD pricing, Canadian shipping and holiday cut-offs explicit in text and feeds.
United Kingdom: consumer referees. which.co.uk (11.8%) and moneysavingexpert.com (10.8%) are among the most-cited sites, and consumer organizations carry 6.5% of citations. 50.0% of timing answers say buy on Black Friday. Local play: make sure your Cyber Monday claims hold up to consumer-body scrutiny, with a real price history.
Germany: Cyber Week, idealo and MediaMarkt. 48.9% of answers frame the sales as Cyber Week, 59.4% of timing answers say buy on Black Friday, and none of 32 recommend Cyber Monday. mediamarkt.de (14.0%) and idealo.de (11.3%) lead citations. 96.2% of “buy it for me” answers decline. Local play: lead with the week, not the Monday, and keep idealo listings and feed prices exact.
France: the strongest deal checks. 54.8% of deal-check answers give specific evidence, 64.9% of answers state a price, and lesnumeriques.com (10.4%) and idealo.fr (8.1%) are top sources. 40.6% of timing answers say it depends. Local play: get accurate specs and prices to French reviewers and comparison sites, and publish warranty and returns terms clearly.
Türkiye: its own November. 44.6% of answers mention Efsane Cuma, 11.11 or Kasım sales, while Cyber Week appears in 1.7%. Trendyol (29.6%) and Hepsiburada (26.7%) lead retailers, and akakce.com is cited in 18.3% of citing answers. 21.9% of deal-check answers skip the question. Local play: tie your offer to local events by name, and keep Akakçe and marketplace listings accurate.
The playbook: 8 plays for brands
Opttab helps 400+ brands get found and recommended by AI answer engines. Each play below comes from the findings above.
1. Give AI a reason to wait, or tell it not to. Evidence: 38.5% of timing answers say buy on Black Friday; only 7.3% say wait for Cyber Monday. Do this: state plainly whether your Cyber Monday offers differ from Black Friday: which products, which prices, which dates (27 November to 30 November). If the offer runs all weekend, say so. In Opttab: Content Studio builds briefs from the tracked timing prompts where you lose, and Answer Engine Analytics shows what AI advises in each market.
2. Make your Cyber Monday deal provable. Evidence: only 38.7% of Cyber Monday deal-check answers give specific price evidence (Black Friday study: 74.2%); Canada is at 15.6%. Do this: publish price history and comparison-friendly data in plain text. Avoid “was” prices you can’t support. In Opttab: Shopping Insights shows which of your products AI surfaces, and Answer Engine Analytics tracks price-validity prompts per market.
3. Name the event the way each market does. Evidence: 37.6% of UK, German and French answers frame the sales as Cyber Week (Germany 48.9%); in Türkiye, 44.6% mention Efsane Cuma, 11.11 or Kasım sales and 1.7% mention Cyber Week. Do this: use local event names in titles, headings and feeds: Cyber Week in Europe, Efsane Cuma and Kasım campaigns in Türkiye, Cyber Monday in North America. In Opttab: run the same prompts by country and language in Answer Engine Analytics.
4. Publish Christmas delivery cut-offs. Evidence: only 3.4% of answers mention Christmas delivery, and 25.1% of delivery and returns answers; AI Overview appears on 97.9% of delivery and returns searches. Do this: put Christmas delivery cut-offs, holiday return extensions and returns windows on one crawlable page per market. In Opttab: the Agent Readiness checklist flags missing policy fields such as returns and delivery, and GEO/AEO ranks the fixes by impact.
5. Close the Canada gap. Evidence: only 38.3% of retailers named to Canadian shoppers are Canadian, 10.1% of priced Canadian answers use a foreign currency, and only 35.9% of citations come from Canadian sites. Do this: state country storefront, CAD pricing and Canadian shipping in text on every key page and in feeds, and give Canadian publishers and creators the right facts. In Opttab: compare Canadian and US answers side by side in Answer Engine Analytics, and check Canadian product visibility in Shopping Insights.
6. Win the referees, market by market. Evidence: Reddit, Facebook and YouTube are the top three cited sites in the US and Canada; idealo.de is cited in 11.3% of German citing answers and akakce.com in 18.3% of Turkish ones; which.co.uk appears in 11.8% of UK citing answers. Do this: be present and accurate in the community threads and videos North American answers draw on, and keep idealo and Akakçe listings aligned with your site prices. In Opttab: see the sources behind each answer, per prompt and market, in Answer Engine Analytics.
7. Be the merchant an agent can hand off to. Evidence: 82.1% of “buy it for me” answers say they can’t buy (Black Friday study: 64.4%); a strict keyword check finds 1 of 162 offering cart or checkout. Do this: connect feeds, stock and checkout so AI hand-offs land on a working path, and test that path end to end (feed, product, cart). In Opttab: build a Brand Agent in Command Center, connected to your store and shipping tools, and publish it through Agent Distribution as a ChatGPT app, a Gemini extension or an agent on your own site. Website Analytics shows which AI agents visit your site and which pages they read.
8. Track many answers, not one rank. Evidence: two answers to the same question share a top pick only 28.7% of the time, and 10.3% of answer pairs share no brands at all. Do this: measure your share across surfaces and markets, prompt by prompt, rather than chasing one position. In Opttab: Answer Engine Analytics tracks visibility prompt by prompt across AI platforms and markets.
The run-up to Cyber Monday (Monday 30 November 2026)
When | Focus | What to do |
|---|---|---|
Now | Audit and content | Run an Agent Readiness check. Confirm AI crawlers are allowed and key pages render without JavaScript. Track timing, deal-check and delivery prompts in every market and language. Publish your Cyber Monday offer story (what differs from Black Friday, which dates, which products) and your Christmas delivery cut-offs and returns on one page. Use local event names (Cyber Week, Efsane Cuma, Kasım). Fix Canadian currency, storefront and shipping in text. |
Before Thanksgiving (26 Nov) | Referees and feeds | Align price comparison listings and merchant feeds with site prices. Share accurate facts with publishers, reviewers and creators. Connect your Brand Agent and test its checkout hand-off. |
Black Friday (27 Nov) | Go live | Publish real prices as soon as they’re live and update feeds the same day. Make clear which offers continue to Cyber Monday. Watch AI traffic and the answers on your tracked prompts. |
Cyber weekend (28 to 29 Nov) | Stay in sync | Keep prices, stock and feeds in sync. Push Cyber Week and Cyber Monday pages, and refresh Christmas delivery dates. |
Cyber Monday (30 Nov) | Monitor | Keep prices and stock in sync through the day. Monitor answers and agent visits, and note which pages AI reads. |
Methodology
Prompts. 60 shopper questions per country, mirrored across six countries in the local language, drawn from real People Also Ask, forum and autocomplete questions. Eight intents: deal discovery (48 searches), timing, Black Friday or Cyber Monday (48), price and deal validity (48), delivery and returns (48), product comparison (42), gift finding (42), retailer choice (42) and delegated purchase, “buy it for me” (42). Three budget tiers. 252 prompts name Cyber Monday and 108 name Cyber Week.
Surfaces. Four AI surfaces: ChatGPT, Gemini, Google AI Overview and Google AI Mode. Copilot was attempted but returned errors for all requests on collection day and is excluded. This report describes the shopper’s experience across AI. It is not a ranking of models.
Collection. Logged-out sessions, geo-targeted to each country, collected on 4 November 2026 through a third-party collection API.
Answers analyzed. 360 searches × 4 surfaces = 1,440 answer slots. Google showed an AI Overview on 329 of 360 searches, so 1,409 answers returned text. Per-answer percentages use these 1,409 answers.
Coding. Each answer was coded against a written codebook (timing advice, deal evidence, retailers and whether they are home-grown, brands, prices and currency, event names and dates, Christmas delivery, purchase behavior, policy detail), with assisted extraction.
Quality checks. An independent second coding pass re-coded a stratified random sample of 144 answers (six per country and surface). Agreement with the main pass: 84.7% for deal evidence (kappa 0.76), 86.8% for purchase behavior (kappa 0.81), 97.9% for timing advice (kappa 0.91), 94.4% for stated prices (kappa 0.88), and 100.0% for Cyber Monday dates, Christmas delivery and refusals to buy.
Citations. 5,602 answer-domain citation pairs across 2,540 distinct domains, classified by type and home market. 2,831 retailer mentions classified by storefront and headquarters.
Statistics. 95% Wilson intervals; cluster bootstrap by prompt for headline figures.
Black Friday comparison. That study was collected on 31 October 2026 and covered 600 prompts on five surfaces, including Copilot, with 2,898 answers. Prompts, surfaces and dates differ, so every comparison here is directional.
Limitations
Logged-out only. Signed-in, personalized or connected-agent experiences may behave differently, especially for buying.
One day, one run. All data come from 4 November 2026, before most Cyber Monday prices were live. Answers vary from run to run, so treat figures as a snapshot, not a trend.
Not like-for-like with Black Friday. The two studies differ in prompts, number of surfaces and timing. Shifts such as weaker deal checks or more refusals are directional signals, not measured changes.
Assisted coding. Codes were checked through a second pass on a sample, not on every answer. Borderline cases, such as generic versus specific deal evidence or “depends” versus a clear timing recommendation, are the most likely to differ.
Retailer classification. “Home-grown” (headquartered in the country) and storefront classification followed written rules and were not checked against a registry.
Missing AI Overviews. Google showed an AI Overview on 329 of 360 searches. The other searches have no AI Overview answer.
What “offers to act” means. Level 3 of the agent ladder includes offers to track prices or set alerts, not only checkout. The strict cart and checkout figure is 1 of 162 answers.
Small groups. Country cells for timing, deal-check, delivery and delegated purchase prompts hold 26 to 32 answers each. Read them as indicative.
Observational. We describe what answers say, not what drives them. The plays are our recommendations based on these observations, not tested interventions.
