eRetailers are building their own AI assistants, and brands have to win inside them

The real shift in AI shopping is not that people use chatbots to research products. It's that retailers are building AI assistants inside their own shops, deciding what those assistants can see, and keeping rival AI out. For brands selling through those retailers, that changes where you compete and what you can measure.

That shoppers now use AI to research products is established. Around half of European consumers already use AI to learn about a category and compare options, according to McKinsey, with usage higher in markets such as Italy and the UK. The more useful question for brands is where that AI actually sits, and increasingly the answer is inside the retailer.

Amazon, Tesco, Walmart and Zalando are building assistants into their own environments, trained on their own catalogues and governed by their own rules. These are not general chatbots a shopper visits and then leaves. They are part of the shop, sitting between the shopper and the product at the point where the decision gets made.

The assistant now lives inside the retailer

Amazon's Alexa for Shopping, the assistant it used to call Rufus, is the furthest along. Amazon says more than 300 million customers used it during 2025, and that shoppers who use it are around 60% more likely to buy. It answers natural-language questions inside the Amazon app and site, drawing on Amazon's own catalogue, reviews and data, and it is live in the US, UK, Germany, France, Italy, Spain, Canada and India.

Other retailers are building the same thing. Tesco has been trialling an in-app assistant that helps shoppers plan meals and fill a basket. Zalando's assistant grew from six million users across the whole of 2025 to around ten million in the first quarter of 2026 alone. Walmart has built its own assistant, Sparky, and even where it appears inside ChatGPT it keeps the catalogue, the basket and the checkout on Walmart's systems.

Amazon has gone further and started selling the technology behind Alexa for Shopping to other retailers that want their own version. The direction is consistent across all of them. Each major retailer wants an assistant it owns, inside an environment it controls.

And they are keeping rival AI out

The other half of the story is what these retailers will not allow. Amazon has blocked dozens of outside AI agents from operating on its site, including OpenAI's ChatGPT, whilst promoting its own tools. In November 2025 it sued Perplexity over the shopping agent in its Comet browser, and in March 2026 a court granted Amazon a preliminary injunction stopping that agent from reaching Amazon's password-protected systems.

Amazon has also stayed out of the open shopping protocols that OpenAI and Google have built, choosing to keep its AI, its data and its advertising inside its own walls. Andy Jassy, Amazon’s CEO, has said agentic commerce could be good for eCommerce, but that outside agents are not yet good enough on personalisation and pricing, and that any partnership would be on Amazon's terms.

The pattern is playing out across the industry. Each major retailer is becoming its own AI environment, with its own rules about what its assistant sees and who else gets in. A general assistant cannot reliably carry your brand into Amazon, because Amazon has shut those assistants out.

What this changes for brands 

The contest moves inside each retailer, on that retailer's terms. Showing up well in a general chatbot does not help you on Amazon if Amazon's own assistant is the one making the recommendation, and a different retailer's assistant will read your products differently again.
It also makes visibility harder. When a retailer's assistant recommends you, or leaves you out, the decision happens inside an environment you do not control and cannot easily observe. You can see the sales, but not the reason behind them, at the exact moment a shopper is forming a view.
There is an execution side to this too. Even as buying through AI matures, retailers want to own that step, which is why Walmart keeps checkout on its own systems. The influence over the decision is already here, so the brands that adapt first will be the ones that treat each retailer's AI as a place they need to be visible and measurable in now.

What good looks like today

The fundamentals matter.. A retailer's assistant will not put forward a product that is out of stock, has lost the Buy Box, is poorly rated or has thin content. Availability, pricing, ratings and content are the price of entry.

Content does the next bit of work. These assistants answer real shopper questions, so product pages that address those questions directly and set out the trade-offs clearly give the assistant something solid to use. On Amazon, Alexa for Shopping draws on listings, reviews, the questions and answers on a page, and the structured attributes behind it, and it favours clear writing over pages stuffed with keywords.

Then there is measurement. Tracking how often, and for which customer needs, your products appear in a retailer's assistant turns a hidden process into something you can manage, including where competitors are showing up more than you and where the gaps are.

What we have built for Alexa for Shopping

Measuring how brands and products show up inside retailer AI is a core part of the eStore platform roadmap, and the first capability is already live. It measures your brand's share of voice inside Amazon's Alexa for Shopping, so you can see how often the assistant puts your products forward and where it does not.

It starts from the customer needs that matter in your category. We analyse the questions shoppers actually ask, drawn from review and category data, frame the questions a shopper would put to the assistant, and then monitor where your brand and products appear in its answers. That turns a process you cannot normally observe into something you can measure and act on.

In practice it shows you four things:

• Share of voice by customer need. How often your brand and products are recommended, broken down by the specific shopper needs and questions that drive your category, rather than a single overall figure.
• Competitor benchmarking. Where you are over- or under-indexing against competitors, and which competitors dominate the questions that matter most.
• Product-level visibility. Which of your SKUs the assistant surfaces, and which it overlooks.
• Content diagnostics. How your product pages read against the questions the assistant answers and against stronger competitor pages, with specific suggestions for titles, descriptions, bullets and specifications.

We are precise about where AI does the work and where it does not. The measurement monitors and reports where you appear. The content analysis uses AI to assess how well a page answers shopper intent. Neither is presented as a black box, because our clients ask us to be clear about it.

Where this goes next

The capability is live for Amazon's Alexa for Shopping today, which is the most established retailer assistant and the one where the data is richest. Other retailer environments will follow as their own assistants mature, and deeper measures of agentic share of shelf, along with tracking the impact of content changes over time, are on the roadmap.

We are early in this and building it carefully, with our focus on the AI search that sits inside the retailers where your products are sold rather than the general chatbots those retailers are keeping at arm's length. The space is moving quickly, so we would rather show you what works today than talk about what might come later.

If agentic search is on your radar, we are glad to talk it through and show you what the current capability turns up for your brand and category. A fuller walkthrough will follow in the coming months.

Sources

• McKinsey & Company, "Europe's agentic commerce moment: decision influence is here; execution is coming", March 2026. https://www.mckinsey.com/capabilities/quantumblack/our-insights/europes-agentic-commerce-moment-decision-influence-is-here-execution-is-coming
• Amazon, "How customers are making more informed shopping decisions with Rufus". https://www.aboutamazon.com/news/retail/how-to-use-amazon-rufus
• Modern Retail, "Amazon says its AI shopping assistant is gaining traction, with Rufus users up 115%", April 2026 (300 million users and incremental sales figures). https://www.modernretail.co/technology/amazon-says-its-ai-shopping-assistant-is-gaining-traction-with-rufus-users-up-115/
• Fortune, "Amazon says its AI shopping assistant Rufus is so effective it's on pace to pull in an extra $10 billion in sales", November 2025 (shoppers around 60% more likely to buy). https://finance.yahoo.com/news/amazon-says-ai-shopping-assistant-152500992.html
• Amalytix, "Alexa for Shopping (formerly Amazon Rufus) 2026" (live markets, and how the assistant reads product content). https://www.amalytix.com/en/knowledge/ai/amazon-rufus-guide-2026/
• Zalando, "Zalando brings its AI-powered assistant to all markets". https://corporate.zalando.com/en/technology/zalando-brings-its-ai-powered-assistant-all-markets-and-adds-four-new-cities-its-trend
• Sporting Goods Intelligence Europe, "Shoptalk Europe 2026 wrap-up: AI and agentic commerce", June 2026 (Zalando assistant user growth). https://www.sgieurope.com/trade-shows/shoptalk-europe-2026-wrap-up-ai-and-agentic-commerce/121631.article
• Tesco PLC, "Tesco launches large-scale colleague trial of new AI assistant", April 2026. https://www.tescoplc.com/tesco-launches-large-scale-colleague-trial-of-new-ai-assistant-designed-to-lighten-the-load-for-customers/
• The Grocer, "Tesco launches meal-planning, basket-building in-app AI assistant", April 2026. https://www.thegrocer.co.uk/news/tesco-launches-meal-planning-basket-building-in-app-ai-assistant/717474.article
• Retail Dive, "Walmart brings Sparky to ChatGPT as OpenAI rethinks Instant Checkout", March 2026. https://www.retaildive.com/news/walmart-sparky-chatgpt-instant-checkout/815647/
• CNBC, "OpenAI's first try at agentic shopping stumbled. It's trying again", March 2026. https://www.cnbc.com/2026/03/20/open-ai-agentic-shopping-etsy-shopify-walmart-amazon.html
• AWS, "Agentic Shopping Assistant on AWS", May 2026. https://www.aboutamazon.com/news/aws/aws-agentic-shopping-assistant-retailers
• CNBC, "Amazon wins court order to block Perplexity's AI shopping agent", March 2026. https://www.cnbc.com/2026/03/10/amazon-wins-court-order-to-block-perplexitys-ai-shopping-agent.html
• GeekWire, "Judge blocks Perplexity's AI bot from shopping on Amazon in early test of agentic commerce", March 2026 (includes Andy Jassy comments). https://www.geekwire.com/2026/judge-blocks-perplexitys-ai-bot-from-shopping-on-amazon-in-early-test-of-agentic-commerce/
• Bloomberg, "Amazon wins court order blocking Perplexity's AI shopping bots", March 2026. https://www.bloomberg.com/news/articles/2026-03-10/amazon-wins-court-order-blocking-perplexity-s-ai-shopping-bots
• Opascope, "AI Shopping Assistant Guide 2026: Agentic Commerce Protocols" (Amazon's absence from the OpenAI and Google protocols). https://opascope.com/insights/ai-shopping-assistant-guide-2026-agentic-commerce-protocols/

References and Further Reading

• The Portman Group - UK alcohol industry self-regulation and labelling guidance
• Drinkaware - responsible drinking guidance and consumer research
• IWSR Drinks Market Analysis - European alcohol ecommerce and no/low forecasts
• Rabobank Beverage Research - online alcohol marketplace and product page analysis
  

David Halls, VP of Sales 

June 2026