What Retail Leaders Are Actually Saying About AI, Trust, and the Digital Shelf at Shoptalk Europe 2026

We spent last week in Barcelona at Shoptalk Europe. Here's what we heard from Douglas, Mondelez, Boots, Arc'teryx, Vuori, and others about where retail is headed.

Shoptalk Europe this year had a theme that kept surfacing across almost every session and conversation: the customer experience is the competitive moat, and AI is reshaping how brands build and protect it. Not hypothetically. Right now.

What stood out wasn’t the AI hype. There’s plenty of that everywhere. What stood out was how specific retail leaders are getting about what actually needs to change in their organizations. The conversations were less about “should we use AI?” and more about “we’re already in it, so how do we operate differently?”

Here’s what resonated most from the week.

Multiple sessions referenced what several speakers called the “agentic era.” Not as a prediction. As a working assumption underneath most of the strategic conversations on stage and on the floor.

Amish Mehta from Boots put it directly: “We are already in a GEO era.” His session on agentic commerce focused on how brands need to engage AI agents to make the shopper experience seamless. Not just optimize for traditional search. Actually redesign how your organization operates to account for the fact that agents are becoming intermediaries between your brand and your customer.

That framing shifts the conversation from marketing tactics to operating model. Mehta talked about matching content to customer journeys, identifying change agents inside the organization, and the opportunity to drive narrative across LLMs and on platforms like Reddit by empowering employees to be active in forums and shape how the brand shows up.

Douglas shared that they’re already seeing AI traffic grow from 4% to 5% of total, and the trajectory is clear. More interesting was their observation about what performs in AI results: reviewers and “expert” content are gaining trust and visibility. Product data alone isn’t enough. You need content that shares positive experiences and builds a believable story paired with findability.

Their take on generational trust was interesting too. The younger the consumer, the more they distrust AI-generated recommendations, which means reviewers and influencers become even more impactful as trust signals. Brands need to think about online and offline visibility together, and the content strategy has to account for both.

Celsius Borges, Head of Digital Commerce at Mondelez, gave one of the more grounded sessions on what GEO actually looks like inside a major CPG company.

His core point: brand is still important, but the shopper has changed. Brand marketers are now responsible for marketing to consumers and to agents alike. That’s a real shift in the job description, and most teams haven’t caught up to it yet.

What I appreciated was the specificity. He talked about measurement foundations: sentiment, citations, and recommendations as the metrics that matter for GEO. He described a holistic story across 30 touch points, where occasions and context drive the narrative. Not “buy Oreos” but “kids afternoon snacks,” the kind of contextual framing that AI models actually pick up and use in responses.

Mondelez is investing in long-form writing and video specifically to give LLMs richer source material to reference. 35% of their attribution comes from earned media. Retailer PDPs are the second largest slice. And trusted third-party sources are, in his words, “even more important” than brand-owned content for building the kind of credibility that AI models rely on.

He was candid about the challenges too. There’s organizational reticence to fully commit to an AI search strategy because “it will probably change.” His advice: pay attention first to recommendations, take small steps, understand the technical parts, then hire someone to manage GEO full time.

That last point is worth sitting with. Mondelez, one of the largest consumer goods companies in the world, is talking about dedicated full-time agentic marketing headcount. That’s a signal.

We’ve been tracking this exact dynamic in our retail GEO research, where we analyzed 28 major retail and consumer brands and found the same pattern Borges described: brands are being mentioned by AI models constantly but rarely cited. The 36-point gap between mention rate (44%) and citation rate (8%) is exactly what happens when brand awareness exists without the content architecture to convert it into AI search traffic. Everything Mondelez is doing with long-form content and contextual storytelling is aimed at closing that gap.

If there was one word that connected every session, it was trust. But what trust looks like is changing.

Douglas framed it as “data, content, and trust are basics.” Not differentiators. Basics. The table stakes for operating in a world where AI is mediating the relationship between brand and buyer.

Mehta at Boots talked about matching on-site experience to the customer journey, needing technology that can surface a custom experience in real time. The signals that matter most aren’t just search rankings anymore. They’re the signals AI agents use to decide whether to recommend you.

Borges at Mondelez noted that strong organic presence matters, but heavy-handed sponsored content kills trust. There’s a tension between wanting to be visible in AI-generated results and maintaining the authenticity that makes those results credible.

Here’s the bigger conversation though: trust requires consistency, and consistency requires speed. When your AI search results shift because a competitor publishes a better buying guide, or a model updates its source preferences, or a new review site gains authority in your category, you can’t wait six weeks to respond. The content fix might be straightforward. But getting that fix prioritized, reviewed, approved, built, and published across enterprise systems is where most teams stall.

This is the execution gap we’ve been writing about in the context of agentic content infrastructure. The insight layer is getting commoditized fast. Every brand at Shoptalk was talking about measurement, sentiment, citations. The execution layer, actually shipping the changes continuously, is where competitive separation happens.

Ashley Kechter from Vuori made a point that cut through the AI-heavy agenda: people will naturally crave authenticity and relationships with retail brands. AI can help with discoverability, but it won’t replace stores. AI can’t replace experience or emotion. Stores need soul.

Their approach to international expansion backs this up. Open aggressively in specific markets, test, then go broad. Measure through social sentiment, customer insights, and competitor data. Their Jack Draper partnership is a good example: they marketed an Indian Wells influencer campaign to a UK audience, meeting customers where they are rather than where the activation happened.

The operating model framework Kechter shared is worth borrowing: on-the-ground leads paired with corporate leads who set non-negotiables. Product teams that mix local and global assortment. Trust your customer. Empower your leaders. Protect your brand DNA. Build capabilities before store count.

It’s a good reminder that the brands pulling ahead aren’t the ones adopting the most tools. They’re the ones with the clearest identity and the operational muscle to express it consistently across every surface, digital and physical.

The “digital shelf” has been a retail concept for years, but the definition is expanding. It used to mean your presence on e-commerce platforms and search results. Now it includes where and how you show up in AI-generated answers.

Douglas’s framing of “SEO will bring the scale, GEO will bring influence” captures it well. Traditional search is still the volume play. But AI search is where buying decisions are increasingly shaped. The consumer who asks ChatGPT “what’s the best running shoe for beginners?” isn’t browsing. They’re deciding. And the brand that shows up in that answer with a citation, not just a mention, is the one that captures the commercial intent.

Mehta at Boots reinforced this: search and AI come before in-store purchases. The digital experience needs to match the customer journey, and that requires technology that can surface the right content in real time.

Borges at Mondelez made the point that content is more important than ever. Community forms around content, then product follows. That’s a different sequence than most retail marketing orgs are built for.

What connects all of this is the speed question. AI search results are dynamic. 40 to 60 percent of cited sources can change within a single month, based on what we’ve seen in our research. A quarterly content audit isn’t a strategy when the shelf resets every few weeks. The brands that win on the AI shelf are the ones that can detect what changed, create or update the right content, get it approved, and publish it, continuously.

There was a question I kept waiting to hear at Shoptalk that never got asked directly: who does the work?

Every session talked about what needs to change. Publish more expert content. Build contextual narratives. Add structured data. Match content to customer journeys. Monitor across AI models. Invest in long-form writing. All of it is right.

Enterprise marketing teams are already at capacity. The content supply chain is already stretched. The stack is already fragmented. Adding “optimize for AI agents” to the existing workload without changing the operating model means it either gets deprioritized or gets done slowly and inconsistently.

This is where the system of work concept matters. Not another dashboard that tells you what to fix. A layer that actually executes the fixes across whatever CMS, DAM, and content systems you’re already running. Authoring, QA, governance, structured data, publishing. The whole chain from insight to live, running continuously.

That’s what we’re building at Gradial. And based on the conversations I had all week in Barcelona, the industry is ready for it. The brands know what needs to change. The bottleneck isn’t strategy. It’s execution speed.

If I had to distill Shoptalk Europe 2026 into three takeaways:

Trust is built through content, not just through brand. The brands earning trust in AI search are the ones publishing authoritative, question-answering content on their owned domains. Brand awareness gets you mentioned. Content authority gets you cited. The research we’ve done across 28 retail brands confirms this at scale.

The operating model has to change, not just the content strategy. Mondelez is hiring full-time agentic marketing roles. Boots is redesigning how they operate as an organization. Vuori is building capabilities before store count. The structural changes matter more than the tactical ones.

Speed of execution is the real competitive advantage. Every brand at the event can identify what needs to change. The ones that pull ahead will be the ones that can actually ship those changes continuously across their stack.

Barcelona was great. The conversations were sharper and more specific than any conference I’ve been to this year.

If you want to see how your brand shows up in AI-generated answers, request a GEO report.