AI-assisted shopping: how merchants should prepare for change

Larry Colagiovanni, Head of Shopping Product at Microsoft AI, shares strategies merchants should adopt to stay relevant in the age of AI-assisted shopping. 

Have we crossed the Rubicon with AI shopping? Are we currently witnessing early transformation, or has consumer behaviour already shifted permanently?

The Rubicon is behind us. We are already in a world where AI has permanently shifted consumer expectations, even if adoption is still uneven. Shoppers have seen what it feels like to get curated outfit ideas, automated meal plans, or side-by-side product comparisons in seconds. Once you experience that, scrolling through dozens of pages of search results feels outdated.

That said, we are still in the early innings of transformation. Traditional discovery paths – search engines, retailer websites, marketplaces – remain the default starting point for most. But the foundation of consumer behaviour has changed: people now expect AI to help them cut through clutter, save money, and provide more personalised recommendations. 

The transformation is still spreading, but the mindset shift is permanent.

From what you are seeing now, how familiar are merchants with AI-assisted shopping, and which types of merchants are most successful at adapting to it? What are they doing differently?

Familiarity varies widely. Large retailers and digital-native brands are leaning in aggressively. They have teams dedicated to structuring product data, testing AI integrations, and optimising for inclusion in AI-driven recommendations. Smaller merchants, by contrast, often don’t yet know how to prepare.

The merchants adapting best share three traits:

1. Data readiness. They maintain clean, detailed product attributes – beyond size and colour – to include fit, style, occasion, and other descriptive context. This makes their products more “AI-discoverable.”
2. Experimentation across channels and owned properties. They don’t just optimise visibility in third-party assistants; they also deploy AI on their own sites. This includes personalised recommendations, conversational shopping agents, and even multimodal interfaces that blend voice, text, and images to make discovery more engaging.
3. Storytelling. They highlight not only their brand values but also the unique qualities of their products – the craftsmanship, materials, or design details that make items stand out. This richer product-level storytelling helps AI surface their catalogue in ways that feel relevant and distinctive.

In short, those who treat AI as both an ecosystem to optimise for and a tool to enhance their own consumer experiences are pulling ahead.

What’s the biggest challenge brands face when optimising for AI-assisted shopping?

The biggest challenge is loss of control over the customer journey – and the evolution from traditional SEO to AI optimisation.

Historically, brands focused on optimising their own sites and winning at SEO so that Bing or Google would redirect shoppers to them. With AI shopping, that dynamic changes. An intelligent intermediary now decides which products or bundles to recommend, and increasingly, the AI may summarise, compare, or even purchase on the shopper’s behalf without sending them to the retailer’s site.

This is both an opportunity and a risk. Smaller brands can surface more easily than in traditional search, but a hero SKU may never appear if its data isn’t structured well or reviews lag behind competitors. And unlike SEO-driven traffic, there are fewer guaranteed site visits or direct brand touchpoints to build loyalty.

Brands now face a strategic shift. It's about optimising for algorithmic shelves where trust, transparency, and machine-readability determine whether AI assistants surface your products in the first place. It also requires rethinking loyalty: with fewer guaranteed visits to a brand’s own site, merchants must find new ways to build ongoing relationships, whether through data sharing, post-purchase experiences, or AI-enabled personalisation that keeps their brand top of mind even when the transaction happens elsewhere.

For smaller merchants who might feel overwhelmed by the transition to AI-driven customer behaviours, what are three actions you believe they should take today to prepare for AI-assisted shopping?

I’d recommend three very pragmatic steps:

1. Clean up product data. AI systems rely on rich context. Merchants should ensure their product descriptions, photos, and metadata are complete and accurate. Even adding detail about fabric feel, use cases, or distinctive features can make a product more likely to surface.
2. Use marketplace AI tools. Instead of reinventing the wheel, lean on Bing, Google, Shopify, Amazon, etc. for recommendations. These platforms are already investing in AI, and small merchants can “borrow scale” from them.
3. Differentiate with narrative. AI will often surface functional options first. Merchants who weave their product stories into pages – whether about craftsmanship, origins, or standout features – give AI assistants richer material to work with and a reason to elevate them. Just as importantly, this mirrors how people are now asking: instead of searching for rigid attributes like size or colour, they frame needs in context (e.g. “a dress for a cocktail party” or “boots for rainy hikes”). Merchants that tell stories at this level of intent give AI systems the context needed to match real consumer queries.

These steps can help future-proof smaller merchants for a world where AI is the first shopping touchpoint.

When it comes to demographics, are there differences in how consumers adopt AI shopping tools?

Yes – adoption isn’t uniform. We’re seeing three main divides:

• Age: Gen Z and Millennials are adopting AI shopping fastest, especially in lifestyle categories like fashion, beauty, gaming, and wellness. They’re comfortable experimenting and value inspiration. Older consumers are slower but adopt AI where it solves practical problems, such as meal planning or grocery savings.
• Income: higher-income households adopt AI earlier because of exposure to premium platforms and devices. But cost-sensitive households are also adopting quickly when AI demonstrably saves money – like surfacing coupons or price-matching.
• Geography: markets like China and South Korea, where super-apps and conversational commerce are already mainstream, are further ahead. The US and Europe are catching up now that AI is being embedded into search, browsers, and messaging apps.

So, while AI shopping will eventually be universal, the on-ramps differ depending on age, income, and regional context.

What metrics should brands be tracking to understand their performance in AI-assisted shopping experiences?

Traditional ecommerce metrics aren’t enough. Brands need new lenses for AI-mediated discovery. A few to focus on:

1. Share of AI shelf. Track how often your products appear in AI responses relative to competitors – essentially your visibility across relevant queries.
2. Assisted conversions. Many purchases influenced by AI will finalise on other platforms. Brands need attribution models that recognise AI as a driver even when it’s not the point of sale.
3. Time-to-clarity. How quickly does an AI system surface your product as the answer to a consumer’s query? This reflects how well your data is structured.
4. Price competitiveness. Because AI tends to highlight value, brands should monitor how often their products surface as the “best deal” in recommendations.
5. Engagement quality. Repeat purchases, reviews, and NPS after AI-mediated interactions reveal whether consumers trust AI-recommended products.

These metrics help brands move beyond measuring clicks to measuring relevance, trust, and competitive standing in the AI ecosystem.

Larry will be speaking at eCommerce Expo, where he will further explore the AI-driven transformation of consumer shopping. eCommerce Expo is the ultimate destination for every B2C and B2B selling online in the UK. Explore ecommerce – from acquisition to logistics – on 24-25 September 2025 at Excel London, with two days of innovative sessions, networking, and solution discovery. Find out more about the event and secure your pass today.

About the author

 

Larry is a leader in the intersection of AI and commerce, currently heading Microsoft AI’s Shopping Products, including Copilot. With over 15 years of experience across roles at Microsoft, eBay, startup Decide, and incubator Madrona Venture Labs, Larry blends technical vision and commercial strategy to help brands and consumers thrive in the rapidly evolving world of AI-driven commerce.

 

About the company

Microsoft creates platforms and tools powered by AI to deliver innovative solutions that meet the evolving needs of our customers. The technology company is committed to making AI available broadly and doing so responsibly, with a mission to empower every person and every organisation on the planet to achieve more.