Barclays says 55% of Black Friday shoppers intend to use AI tools. KPMG finds 12% of consumers plan chatbots. What should UK beauty and fashion retailers actually fix?
The shopper arrived with an assistant. The product page arrived with vibes.
The shopper's new digital assistant doesn't care how expensive your homepage looks if it cannot tell a 30ml perfume from a 75ml one. Imagine an impeccably art-directed fragrance campaign leading to a page that forgets the bottle size, delivery deadline and return policy. The assistant cannot answer. Neither can the human customer. That is an impressively expensive way to lose a sale.
Fresh Barclays research reported on 17 November 2026 says 55% of Black Friday shoppers plan to use AI tools to help find deals, rising to 67% for Gen Z shoppers. The bank forecasts a £10.5bn sales period and reports that 46% of UK adults intend to take part (Barclays, 2026). Those figures are survey intentions and forecasts, not measured spending driven by AI. Black Friday is 27 November; Cyber Monday follows on 30 November.
55% says Barclays. 12% says KPMG. Both deserve context.
KPMG's 10 November poll of 3,000 people found that 12% of UK consumers plan to use AI chatbots such as ChatGPT or Gemini to search for Black Friday offers, down from 15% the year before. Another 13% intended to follow AI-suggested search links. The most common route remained retailers' own websites and apps (KPMG, 2026).
That is not a direct contradiction. Barclays refers to a broad range of AI tools among people planning to shop the sales; KPMG's chatbot figure has a narrower definition and a broader consumer population. Different wording, samples and dates make the percentages unsuitable for a simple head-to-head comparison.
YouGov's October study offers a third lens: 34% of likely Black Friday and Cyber Monday shoppers planned to use AI, compared with 19% the previous year (YouGov, 2026). Again, a different research question. None establishes that half the country will buy something because a chatbot recommended it. The responsible conclusion is that assisted comparison is becoming relevant to a meaningful group of shoppers, while retailers and ordinary search remain essential.
Price isn't just a number. It's a promise.
Barclays' findings suggest more shoppers are checking value and reviews before buying. YouGov found that 47% of likely sales shoppers struggle to know whether discounts are genuine (YouGov, 2026). A dramatic percentage-off banner is a poor substitute for evidence.
The ASA's October 2026 advice requires savings claims to be genuine and not misleading. It says a reference price should reflect a genuinely established usual selling price, with pricing history, duration and channel among the factors examined (ASA/CAP, 2026). The CMA also publishes guidance on unfair commercial practices under the Digital Markets, Competition and Consumers Act (CMA, 2025).
For a beauty gift set, explain exactly what is included, the size of each product, the real basis of the claimed saving and any exclusions. For a dress or coat, make clear which variants are reduced and when the offer expires. Save your price-history evidence. A customer who can verify the offer is more useful than a shopper briefly impressed by a strikethrough.
Your catalogue should be ready for scrutiny, not just screenshots.
Begin with truthful product attributes. Fashion needs accurate composition, colour, size availability, fit guidance, delivery and returns. Fragrance needs clear volume and concentration where relevant. Skincare needs legible ingredients and carefully substantiated claims. A shopper may begin with an occasion or a budget rather than your catalogue's exact product name.
Google recommends both Product structured data and Merchant Center feeds to help its systems understand products, price and availability. A feed also gives retailers greater control over updates, though neither approach guarantees visibility in Search or a recommendation by an unrelated AI tool (Google Search Central, n.d.).
If the website says £28 but a shopping feed says £35, or a bestselling shade is sold out but still marked available, the experience is broken before the campaign gets a chance to work. Fix the source catalogue and synchronisation first. Don't buy a glamorous new chatbot to sit on top of inaccurate stock.
Don't trade beauty for data. Design them together.
An excellent product page can still feel like a fashion editorial. Use beautiful photographs, moving creative, considered typography and a clear point of view. Then provide readable practical information where customers expect it: visible text, understandable filters, a useful size guide and accessible mobile navigation.
Show the dress on a person and identify the displayed size. Explain the handbag's measurements and material. For a scent, let the campaign convey the fantasy while the page explains notes, volume, gifting and delivery. The point isn't to make every shopper read a spreadsheet. It's to remove the uncertainty standing between desire and purchase.
Think in real customer questions. What perfume can I give someone who dislikes sweet scents? Which jacket works over a wedding outfit? Will this serum be appropriate for my routine? Where a claim needs specialist substantiation, do not invent an answer. Useful content should connect a question to evidence, not produce endless generic SEO pages.
The 14-day readiness test for a fashion or beauty retailer.
Days 1–3: select 20 high-priority products. Audit each mobile page for missing details, inaccurate stock, price discrepancies, unhelpful variants, confusing delivery information and broken purchase steps. Record current sessions, product-page conversion, basket additions and customer-service questions as a baseline.
Days 4–6: correct the catalogue and any linked shopping feeds. Make prices, currency, availability, sizes, variants, promotional dates and delivery terms consistent. Confirm meaningful Product structured data where applicable and make important facts available in rendered page text, not locked in an image.
Days 7–10: publish a genuinely helpful comparison or buying guide for an important product family. A fragrance gifting guide could explain how to choose without pretending scent is universal. Fashion can help readers decide by fit, material and occasion. Use original or properly licensed images and make the next step to a suitable product obvious.
Days 11–14: test common questions using internal site search, ordinary Google results and several shopping assistants, recording wrong, missing or outdated answers. Don't treat one chatbot output as a stable ranking. Compare changes against your baseline while accounting for promotions, paid media, stock and product mix.
Measure a better purchase, not a mystical visibility score.
Look at qualified sessions, conversions, margin, returns, failed searches, customer support contacts and what customers say was difficult. Search Console can show Google discovery; it cannot reliably allocate every purchase to every private AI conversation. A survey about planned use doesn't tell you whether your own retailer has gained incremental revenue.
Be sceptical of anyone selling a definitive percentage of all AI recommendations without a transparent methodology. A modest audit of real product questions, checked regularly, is more informative than a big number in a fashionable dashboard.
The machines can be wrong. So can the marketing department.
AI tools can repeat old prices, misunderstand a query or surface unauthorised resellers. YouGov reports that 48% of AI shoppers are worried about biased or sponsored recommendations (YouGov, 2026). Keep your official domain, customer support and verification details easy to find. Review claims and price comparisons rather than assuming software will catch every error.
Never optimise so aggressively for automated reading that you forget people who prefer a shop floor, a trusted retailer or a straightforward checkout. KPMG's findings are a useful reminder that plenty of customers still arrive directly (KPMG, 2026). A good digital experience should welcome both kinds of buyer.
Questions smart brand teams should ask.
Does 55% mean more than half of UK adults will shop through ChatGPT? No. Barclays reports intended use of broadly defined AI tools among Black Friday shoppers, not chatbot use by 55% of all adults. KPMG's 12% figure refers more specifically to consumers intending chatbot deal searches.
Will adding Product schema guarantee an AI recommendation? No. Google documents eligibility and product understanding on Google's own surfaces, not a universal AI ranking system. Accuracy and consistency are worthwhile without an invented guarantee.
What should a smaller beauty brand do first? Improve sale-price evidence, size and product facts, delivery clarity, mobile purchasing and trustworthy creative. Then test how actual customers discover and compare its offer.
A brilliant campaign makes them look. A brilliant shop gives them a reason to stay.
Whether the final AI-shopping number is closer to 12%, 34% or 55% depends on whom you ask, and what counts as AI. Sensational certainty would make a poor conclusion to a story about trust. The better question is simpler: can your brand give a customer a complete, accurate and desirable reason to choose it?
The Digital Dollhouse is a Knutsford-based digital marketing and creative technology agency. We bring human creative direction, commercial strategy, beautiful design and robust websites together. Explore our services, our portfolio or talk to us about a better ecommerce journey. For more on turning seasonal attention into a sale, see our Winter Edit.
References
Barclays (2026): Black Friday 2026 consumer-spend research, 17 November (Barclays press material reproduced by Insight DIY) https://www.insightdiy.co.uk/news/barclays-uk-shoppers-set-to-spend-105bn-during-black-friday-sales/17033.htm
KPMG (2026): AI use for Black Friday deal searching stalls, 10 November https://kpmg.com/uk/en/media/press-releases/2026/11/ai-use-for-black-friday-deal-searching-stalls.html
YouGov (2026): The bots and the basket: AI's role in Black Friday and Cyber Monday 2026 in Great Britain, 8 October https://yougov.com/en-gb/reports/55674-uk-mega-sales-report-2026
Google Search Central (n.d.): Share your product data with Google, official documentation https://developers.google.com/search/docs/specialty/ecommerce/share-your-product-data-with-google
ASA/CAP (2026): Promotional savings claims, 6 October https://www.asa.org.uk/advice-online/promotional-savings-claims.html
Competition and Markets Authority (2025): Unfair commercial practices: CMA207, updated 18 November 2025 https://www.gov.uk/government/publications/unfair-commercial-practices-cma207
Bottomley, T. (2026): More than half of Brits plan to use AI tools to find best deals this Black Friday, TheIndustry.fashion, 17 November https://www.theindustry.fashion/more-than-half-of-brits-plan-to-use-ai-tools-to-find-best-deals-this-black-friday/

