For more than a decade, digital commerce followed a familiar playbook: SEO and marketing campaigns drove traffic, UX guided the journey, and a wide selection of on-trend products sealed the deal. The underlying assumption was simple – get a human to the site, and branding, design, and persuasion would do the rest.

That assumption is quietly collapsing. Just a few days ago, Google launched the Universal Commerce Protocol (UCP) – a new open standard designed to let AI agents handle the entire shopping journey: from product discovery to checkout and post-purchase support, removing traditional clicks, SEO funnels, and custom integrations between merchants, agents, and payment systems.

Drawing on real-world retail analytics, this expert column is prepared by Galyna Pustova, CEO of Catomize by hidden hint software development, a Swiss-engineered solution designed to optimize catalog management and improve business conversion rates.

AI isn’t just improving search. It’s changing who searches, how decisions are made, and – crucially – whether a click happens at all. We’re entering a zero-click market, where machines increasingly decide what gets bought, and where outdated or unreliable data can make a retailer invisible before a human ever sees the brand.

From searching to delegating

Think about planning a mountain hike. You don’t just need a jacket – you need to account for temperature ranges, terrain, route difficulty, layering systems, weight, airline restrictions, and what makes more sense to rent versus carry. A generic packing list won’t help. What matters is synthesis: how all these constraints fit together.

Doing that manually – across multiple stores, checking sizes and availability, comparing logistics, and making sure nothing is missing – can easily consume an entire day. This is exactly where AI steps in. I could let it compare, reason, and synthesize. The “search” happens on the AI’s side, not on individual websites.

That shift is already measurable. As AI becomes a new front door to the internet, a significant share of traffic is at risk – not because demand disappears, but because decisions are increasingly made upstream, before a user ever clicks.

From search engines to answer engines

In early 2025, AI-generated overviews began appearing in Google Search results. Their presence in e-commerce queries is still limited – only 4% compared to sectors like education (71%) and healthcare (83%), – but that’s not a signal of irrelevance. It’s a lag indicator.

E-commerce is structurally harder. Prices, inventory, delivery promises, and fulfillment must be verified in real time. AI doesn’t guess. It verifies. And once AI systems trust the underlying data layer, adoption accelerates quickly, forecasted to reach USD 31 billion by 2028.

The broader trajectory is clear. The global AI-in-retail market is on track to reach tens of billions of dollars within a few years. More importantly, the nature of search is changing. Users don’t browse – they delegate. And increasingly, AI answers without sending traffic anywhere.

AI does not shop like a human

Traditional UX assumes emotion, aesthetics, brand affinity, and persuasion. AI doesn’t care about any of that.

AI-driven systems already evaluate retailers across dozens of variables: availability accuracy, pricing consistency, delivery reliability, historical performance, response speed, and data consistency across channels. The deepest shift isn’t recommendation – it’s agency.

AI agents already book flights and accommodation. Voice assistants place orders. Smart devices reorder supplies. Autonomous systems are beginning to finalize purchases with minimal or no human confirmation.

According to Statista, 49% of global CEOs believe machine customers will represent a significant customer base by 2030. This creates a new asymmetry: machines evaluate retailers using rules humans never cared about.

Why reliability now outweighs SEO, UX, and brand affinity

AI doesn’t “experience” your website. It reads source code. It parses structured data. It verifies inventory and delivery promises before recommending a retailer – often without opening the site at all.

That dramatically raises the cost of error. If your ERP, warehouse, website, and marketplace feeds aren’t synchronized, the AI doesn’t hesitate. It excludes you. There’s no emotional forgiveness, no chat support fallback, no “maybe I’ll try anyway.”

For luxury brands especially – where perception historically outweighed technical precision – this shift is existential. AI doesn’t care about storytelling or visual identity. It cares about reliability.

Wrong availability? Unreliable.
Price mismatches? Unreliable.
Slow or inconsistent responses? Unreliable.

And unreliable sources don’t get recommended again.

The real revenue leak isn’t marketing

After my team spent more than 15 years in the technical core of e-commerce businesses across Europe and the U.S., we made one conclusion that consistently contradicts boardroom narratives: in an AI-mediated market, the most expensive revenue losses rarely come from weak marketing.

They come from unreliable data.

For years, marketing inefficiency was a reasonable explanation. Humans explored, compared, hesitated – and forgave imperfections. AI changes the logic entirely. AI doesn’t explore; it verifies. It doesn’t persuade; it filters. And increasingly, it decides which retailers are even eligible to be shown.

By 2028, Gartner predicts that 33% of enterprise software applications will include agentic AI, up from less than 1% in 2024. This includes at least 15% of day-to-day work decisions being made autonomously through agents.

In this environment, issues that once hurt conversion now block selection altogether. Incorrect stock visibility isn’t a checkout problem anymore – it’s a pre-selection penalty. If availability data is inconsistent, AI agents simply route demand elsewhere.

Zero-click becomes decisive

Manual workarounds only accelerate the leak. Scheduled imports, spreadsheets, and human overrides introduce latency and ambiguity – the exact conditions AI systems penalize.

In a zero-click, agent-driven market, AI doesn’t compensate with brand affinity. It doesn’t ask clarifying questions. It doesn’t tolerate “almost correct.” It either trusts your system – or stops querying it.

That’s why revenue leakage in the age of AI agents is no longer primarily a marketing problem. It’s a data integrity and system reliability problem.

The good news, an agent-mediated purchasing can rapidly scale once merchants become “agent-ready.” McKinsey estimates that by 2030, agentic commerce could orchestrate $900B–$1T in US B2C retail revenue and $3T–$5T globally.

From traffic optimization to system integrity

The strategic implication is straightforward: e-commerce is shifting from persuasion to verification.

Winning in a zero-click market means:

• Moving from manual, file-based updates to event-driven systems
• Ensuring every price, stock, and availability change propagates instantly
• Designing infrastructure that remains consistent under peak load
• Optimizing not just for humans, but for machines acting on their behalf

The market without clicks is already here

The zero-click market doesn’t eliminate e-commerce; it eliminates excuses. AI doesn’t negotiate, empathize, or wait. It rewards accuracy, speed, and systemic honesty.

Retailers that continue optimizing banners, creatives, and funnels while neglecting backend truth won’t disappear because customers reject them—but because machines stop seeing them. In the next phase of e-commerce, success won’t belong to the loudest brands. It will belong to the most reliable systems.

Speed, accuracy, and trust are no longer technical metrics. They are the new currencies of retail.

Featured image credit