AI’s Impact on Holiday Shopping

Subscribe to Tech Mode by Marc Bain to go deep on the most intriguing developments in AI and how technology is reshaping the fashion industry.

Dear BoF Community,

Welcome back to Tech Mode, your monthly guide to how AI and other technologies are reshaping the fashion industry.

If you’re in the US, I hope you had a nice Thanksgiving, and if you’re not (or don’t celebrate the holiday), I hope you enjoyed the Black Friday and Cyber Monday shopping rush, which has become an international phenomenon. This is the season when retailers are competing their hardest to win customers. In the US this year, they’re vying for consumers whose confidence keeps falling, and yet they’re still shopping.

Both consumers and retailers are increasingly relying on AI during this crucial period. I’ll explain how below, and also unpack the surprise departure of EssilorLuxottica’s wearables chief, explain what the launch of Google’s newest AI model means for fashion and take a look at Alibaba’s plans to make AI shopping agents a thing in China.

But before we jump in, if you didn’t spot it earlier this week, Loewe’s top marketing mind, Charlie Smith, is leaving the company for an unexpected new role: He’s joining London-based tech company Nothing, which makes devices like phones and watches, as its new chief brand officer. Smith said in a statement that Nothing’s mission of making technology fun resonated with him, and that he’s excited to take what he learned in luxury and apply it to consumer technology.

Also, Valentino is the latest fashion label to be called out for using AI in an ad, with criticisms of its surreal video with multidisciplinary artist Total Emotional Awareness including that it looks “cheap” and “not on brand.”

Let’s get to it.

AI Goes Holiday Shopping

ShopGPT: Traffic to US retail sites from AI services soared this Black Friday and Cyber Monday.

• Adobe Analytics said traffic to US retail sites from AI sources was up 805 percent on Black Friday and 670 percent on Cyber Monday compared to last year. For the whole period from Nov. 1 to Dec. 1, AI traffic surged 760 percent year over year.
• Salesforce said AI platforms and AI agents influenced 20 percent of orders globally across the entire Cyber Week, driving $67 billion in sales.

All that growth comes off a small base, but it’s still a big jump and the numbers are only going up. Not only that, Adobe found that shoppers visiting a site through an AI platform on Black Friday were 38 percent more likely to convert than non-AI traffic. The data reinforces what’s become increasingly evident: AI is set to become an inescapable part of online shopping.

Features galore: Tech companies are doing their best to ensure they become a fixture in the shopping journey by introducing new features to make finding products easier. In just the past few weeks:

• OpenAI launched a capability called shopping research that takes the user’s initial query, asks follow-up questions and uses what it knows about the user to build them a personalised shopping guide. It also recently added Target to the list of companies whose products can now be purchased directly in ChatGPT, putting it in the company of Walmart, Etsy and Shopify.

• Perplexity unveiled its own shopping assistant that it says remembers your past searches and learns your patterns. “If you’ve been looking at minimalist running shoes for an upcoming marathon, it prioritizes simplicity when you buy a new bag before race day,” the company said in a blog post.

• Google upgraded the shopping experience in its AI mode to provide information like price, reviews and inventory info when you ask a product question, though by far the most interesting feature lets you ask Google to call nearby stores to find out if an item is in inventory and get details on pricing and promotions. An AI agent will actually ring up the store, and then email or text you with answers, according to the company. (I am very curious what the experience is like for the sales associate answering the call.)

Online shopping’s AI future: The picture that starts to emerge of where shopping is headed is one where finding and buying products online looks very different than it does today. Instead of doing a search and looking at Google’s Shopping section, or opening a bunch of tabs to compare prices and read reviews, you could have a conversation with an AI platform that already knows your preferences. It presents what it thinks are the best products based on your past activity and the information available online, including user reviews and content published about them, and offers to buy them for you or learns where they’re in stock nearby, in your size and preferred colour.

This vision of shopping is unlikely to fully replace how we currently shop online. Sometimes it’s fun to browse a retailer’s website specifically because you don’t know what you’re looking for and want to see what’s new or what the buyers thought was worth stocking. But let’s say half of your searches start to look like this; that’s hundreds of billions in annual sales — or more. McKinsey estimates that, by 2030, agentic commerce could account for $1 trillion in sales in the US, rising as high as $5 trillion globally.

A Key Player in the Ray-Ban Meta Partnership Exits

Ciao, Rocco: Rocco Basilico, chief wearables officer at EssilorLuxottica — and one of the heirs of Luxottica founder Leonardo Del Vecchio — is set to exit the eyewear giant this January. Basilico spent more than a decade at the company, six of them working on its smart glasses with Meta, which have been a success. I talked to him a few times over the years, including just a couple months ago for this interview we included in The BoF-McKinsey State of Fashion 2026.

Basilico seemed dedicated to ensuring that the look of the glasses was never compromised by the technology. When I asked him what the Ray-Ban Meta glasses did to succeed where previous smart glasses had failed, he said, “We started with design first, and we were very rigid in the partnership on not compromising on design.”

“AI glasses wouldn’t be as popular as they are today without you,” Meta’s Mark Zuckerberg said to Basilico in a post on Threads.

Next up: EssilorLuxottica told Reuters that Basilico is exiting to “pursue personal interests,” though it also noted that his departure comes amid some friction within holding company Delfin, which owns a 32 percent stake in EssilorLuxottica. Del Vecchio’s heirs have apparently had some differences over the past three years regarding the succession process. Reuters cited a report from Italian media saying the youngest heir, Leonardo Maria Del Vecchio, had challenged Basilico — the son of Del Vecchio’s wife from a previous marriage, hence the different surname — over his rights in the holding company.

What’s clear is Basilico doesn’t plan to retire young. “Now it’s time for a new chapter,” he posted on Threads, “one that brings together everything I’ve learned: fashion, design, technology, storytelling, and the desire to build something future-facing from the ground up.”

What about EssilorLuxottica: EssilorLuxottica and its investors don’t appear too concerned about Basilico’s departure stifling the company’s big ambitions for smart glasses. EssilorLuxottica told Reuters its talent pipeline in wearables is “very healthy” and that it’s already well-established as the category leader. Analysts at investment bank Equita said in a note that they don’t anticipate any impact on the group.

America’s Next Top Model

Google is back in the game: Google’s latest AI model went public a couple weeks ago and immediately caused a stir. “Holy shit. I’ve used ChatGPT every day for 3 years. Just spent 2 hours on Gemini 3. I’m not going back,” Marc Benioff, Salesforce’s co-founder and chief executive, posted on X. “Call it America’s next top model,” wrote The Wall Street Journal, noting that it surpassed ChatGPT and others on a number of industry benchmarks.

Google had been widely seen as falling behind in the race to build the best LLM, a surprising turn for the company whose research produced the transformer architecture LLMs run on. With Gemini 3, the company has reasserted its AI might.

Immediate impact: The ramifications aren’t just that consumers and companies may opt to use Google’s AI over those of rivals. In an internal memo seen by The Wall Street Journal, OpenAI’s Sam Altman declared a “code red” and instructed staff to focus on improving ChatGPT while shelving other efforts for the moment. Among the projects delayed were AI shopping agents and advertising in ChatGPT, something long rumoured to be in the works.

Winner take all: As much as shopping and ads matter to OpenAI’s plans, they’re secondary considerations. The single most important thing is for ChatGPT to become the default AI platform users turn to for work, school and everything else. As Google itself showed with products like Gmail and Chrome, once you do that, you have a fairly captive audience that you can then advertise to and help with shopping.

OpenAI has a head start in this race, but the tech world has shown time and again that an early lead is no guarantee of long-term dominance. Before Chrome, it seemed like Internet Explorer could never be supplanted as everyone’s preferred web browser, just like Netscape Navigator before it. Myspace was the social network of choice before Facebook. Sam Altman knows ChatGPT’s position is far from unassailable. In five or 10 years, users could be going to Gemini — or some other AI model entirely — to do their research and maybe buy a new pair of jeans.

China’s AI Shopping Agent

Alibaba vies for the AI crown: Tech giants aren’t just trying to change online shopping in the West. China’s largest e-commerce company, Alibaba, will introduce agentic-AI shopping features to its main AI app amid a larger revamp of the app intended to make it more like ChatGPT, Bloomberg reported. The agentic AI would support shopping on platforms like Alibaba’s main Taobao marketplace, possibly as a way to draw more users in as the app tries to make up ground against competitors with greater popularity, such as the app Doubao by TikTok-creator ByteDance.

It’s still unclear exactly what those agentic features might look like, but there’s a good chance they’ll resemble some of the features US-based tech players have explored, like having AI complete checkout on a user’s behalf. The app already had more than 10 million downloads in the week following its relaunch.

Early adopters: China’s online shoppers have historically been quicker to embrace behaviours like social shopping and the use of super apps, where you can do everything from your banking to buying fashion in one place. Despite companies like OpenAI and Google spearheading the latest wave of AI development, it could still be that China is where agentic shopping takes off first.

No guarantees: The obstacles to getting consumers to shop with agents aren’t just about an unwillingness to try new technology. The agents available for shopping can still be slow and clunky, at least in markets like the US and UK. Even Andy Jassy, chief executive of Amazon, which has a shopping chatbot called Rufus, admitted they could be better in an October earnings call. “We have to find a way, though, that makes the customer experience good,” he said. “Right now, I would say the customer experience is not … The delivery estimates are frequently wrong. The prices are often wrong.”

If China can crack agents that actually work well, that could be the real victory.

Want to dive deeper into an insight from this article? Check out The Brain of Fashion, BoF’s new generative AI tool where you can unlock BoF’s technology archive with a single question.