For a decade, business leaders have comforted themselves with a tidy commercial compromise: bricks and clicks. You managed your physical storefronts, optimised your digital search bars, and assumed you had the modern customer journey covered.

But that dual-shelf model just got significantly more complicated, says Rob Gonzalez, Co-Founder and Chief Strategy & Innovation Officer at Salsify, as retailers enter the era of the agentic shelf, the third pillar of the modern shopping journey.

Seamless shopping behaviours redefine the retail shelf

Consumers do not shop in isolated channels anymore; their buying behaviour moves across a fluid loop where physical stores, digital product detail pages (PDPs) and AI-powered conversational assistants coexist within the exact same ten-minute shopping window.

In high-adoption markets, nearly half of consumers (46%) have already integrated AI into their shopping, and one in four now uses an AI agent as their primary starting point, outstripping traditional brand websites for the first time.

European shoppers are actually leading this charge. Recent McKinsey data suggests that 63% of European consumers are actively using AI to compare brands and models during their purchase journey.

But, as a retail leader, you have to look closely at where the machine’s operational influence hits a hard wall. McKinsey’s research highlights a clear, downward slide in consumer confidence the closer the AI assistant gets to the actual “Buy Button” – specifically when it comes to prefilling baskets, completing checkouts or automatically reordering items.

The takeaway here is stark: the AI gets the assist, but your PDP closes the sale. The agent handles the “help me choose” research, but your product data must provide the “help me buy” verification.

The structural trap of siloed data

This reality exposes a structural trap within most traditional retail brands. We are still organised as if these shelves operate in separate universes. Your bricks-and-mortar team focuses on physical constraints, while your digital team spends its retail media budget optimising keywords for a human search bar. Meanwhile, the foundational data required to satisfy the machine concierge is left completely neglected.

An IPG survey of CPG executives found that 90% of online product pages remain completely untouched for months or even years. Leaving your long-tail catalogue out in the cold might have saved headcount costs on the traditional digital shelf, but on the agentic shelf, thin data is an immediate disqualifier.

To stay in the game, brands must extend their PDP “Brilliant Basics” across their entire product line. Most enterprises safely clear this line for their top ten bestsellers on Amazon or Tesco, where the product pages look pristine. But your fiftieth bestseller, or a core item on a tier-three customer site, is likely sitting at a C-minus at best.

When an AI agent scans your entire digital surface area to build a recommendation, it reviews both your best and your worst pages. If it detects inconsistent specs or thin data, the machine flags a logic gap. Because AI models are deeply risk-averse and hate hallucinating, they will exclude your product entirely rather than take a guess.

Winning the opportunity at machine speed

Winning these critical commercial opportunities means shifting from managing simple technical specifications to orchestrating a dynamic context layer across your entire catalogue.

While basic digital hygiene provides the technical facts (the “what”), the context layer provides the “so what?” – explicitly mapping your product attributes to real-world consumer use cases, environments and subjective needs.

Optimising your PDPs everywhere drives immediate traffic and conversion on the digital shelf today, while ensuring you win the AI’s recommendation tomorrow. It allows you to capture the aggregate revenue of the long tail that is currently left sitting on the table.

Now, right about now, you should be deeply sceptical. Manually writing situational context for thousands of legacy SKUs sounds like an operational nightmare that requires doubling your headcount. But you do not solve a machine-age problem with manual labour; you solve it at machine speed.

By leveraging automation and intelligent product experience management (PXM) platforms to handle the heavy lifting of data mapping and enrichment, you protect your bottom-line margins while unlocking massive top-line growth.

The landscape has fundamentally shifted, but it represents a career elevation, not a threat. The brands that survive the next decade of AI-fuelled commerce will be the ones that stop chasing every shiny object and start mastering the brilliant basics across their entire digital footprint.

Your customers are already moving fluidly across all three shelves; it is time to build the engine that meets them there.