Google Is No Longer the Gatekeeper
- Mayte M.G.
- 2 hours ago
- 5 min read
In this article, I'd like to explore how AI agents are rewriting the rules of visibility in ecommerce. Some of these changes are already happening—and some are strategies you should start implementing today if you want to stay competitive.
The Signal Nobody Is Talking About Loudly Enough
In April 2026, Adobe Digital Insights released a data point that should have triggered emergency board meetings across retail. Analysing over one trillion U.S. retail site visits, they found that traffic referred by AI engines now converts 42% better than non-AI traffic. Twelve months earlier, the same dataset showed AI traffic performing 38% worse. That is an 80-point swing in a single year.
Let that number settle for a moment. Because what it means is that the consumer who arrives at your product page through ChatGPT, Perplexity, or Google's AI Overviews is not a casual browser. They arrive already pre-qualified, pre-convinced, and with intent.
The question for every ecommerce manager in 2026 is therefore not 'Should we think about AI search?' It is: 'Why is our brand invisible when an AI answers a category question?'
From SEO to AEO: A Necessary Vocabulary Update
For two decades, Search Engine Optimisation defined how ecommerce teams approached digital visibility. You ranked for keywords. You earned backlinks. You competed for the first ten blue links on a Google results page.
That model is not dead, but it is incomplete. A new discipline has emerged alongside it: Answer Engine Optimisation (AEO). Understanding the distinction is now a professional requirement for anyone managing an online channel.
The measurement panel changes entirely. Traditional SEO is essentially a Google game. AEO is measured across a panel: ChatGPT, Perplexity, Claude, Gemini, Google AI Overviews, and an expanding ecosystem of agent-mediated surfaces. A brand that ranks first on Google but is never cited by an LLM (Large Language Model, the AI systems behind tools like ChatGPT or Gemini) has an increasingly significant blind spot in its visibility strategy.
How the New Discovery Funnel Works
Consider how a real consumer behaved in 2022 versus today.
In 2022: type keywords into Google, scan ten results, visit two or three pages, compare manually, purchase.
In 2026, an increasing segment of shoppers — particularly high-intent, high-spend consumers — behave differently. They open ChatGPT or Perplexity and ask a conversational question: 'What is the best air purifier for a bedroom under 20 square metres, under €200, with a HEPA filter?' The AI responds with a direct recommendation. If your brand is in that answer, you exist. If it is not, you were never in the consideration set.
Two distinct search behaviours are now emerging:
• Quick Search (AI-mediated): Zero-click, AI-answered queries for facts, comparisons, and recommendations. This is where AEO dominates.
• Deep Search (human-driven): High-stakes research where buyers still want to read reviews, verify sources, and compare in detail. Traditional SEO and expert content still win here.
The critical insight: these two behaviours are not competing for the same consumer at the same moment. They serve different stages of the funnel, and a mature digital strategy must address both.
What AI Engines Actually Look For
Here is where many ecommerce teams make a costly assumption: they believe that ranking well on Google automatically means performing well in AI search. It does not. The optimisation logic is fundamentally different.
Traditional SEO rewards lexical keyword matching, domain authority, and backlink equity. AEO operates on semantic entity relationships. An AI model is not scanning for keyword density; it is asking: Does this brand have a clear, extractable identity? Is the data structured so I can confidently synthesise an answer? Are there third-party corroborating sources that confirm what this brand claims?
The five signals that matter most for AEO in ecommerce:
• Entity clarity: Is your brand, product category, and value proposition unambiguously defined in structured data (Schema.org, product schema, review schema)?
• Passage extractability: Can an AI pull a clean, standalone answer from your category pages and product detail pages within the first 500 tokens — without needing to read the whole page?
• Third-party citation: Are independent sources (specialist media, review platforms, comparison sites) naming your brand in the right attribute context?
• Machine-readable signals: Have you implemented llms.txt at your domain root? Is your robots.txt configured to allow AI crawlers?
• Review sentiment at scale: AI models weight authentic customer sentiment. A brand with 4,000 verified reviews averaging 4.6 stars will be cited more confidently than one with a thinner review base.
Three Ecommerce Brands Getting This Right
1. Casper (Sleep & Mattresses)
Casper has restructured its category content around conversational questions rather than keywords. Their mattress comparison pages now open with direct, first-paragraph answers to questions like 'What is the best mattress for side sleepers?' — front-loading the extractable answer before any marketing language. The result: consistent citation in ChatGPT and Perplexity for sleep-related queries, despite operating in a crowded category.
2. Decathlon (Sports Equipment)
Decathlon's European teams have implemented comprehensive product schema across their 80,000+ SKU catalogue, including structured size guides, activity suitability tags, and verified review aggregation. When a user asks an AI assistant 'What running shoe should I buy as a beginner for under €100?', Decathlon products appear in recommended answers at a frequency their competitors — who have not invested in structured data — simply cannot match.
3. Zalando (Fashion)
Zalando has moved beyond product schema to invest in editorial corroboration — commissioning third-party style guides and trend reports that independently cite Zalando brands in context. This creates the off-site citation architecture that LLMs rely on when building confident recommendations. The lesson: AEO is not just an on--site technical exercise. It requires an off-site content ecosystem.
The Practical Playbook: Where to Start
For an ecommerce manager building an AEO capability in 2026, the prioritisation is clear. Start with measurement — you cannot optimise what you cannot see. Build a citation monitoring dashboard across the five major AI engines. Track mention rate and citation rate as separate KPIs alongside your existing SEO metrics.
Then conduct a passage extractability audit on your highest-priority category pages and product detail pages. Ask one question: if an AI model reads only the first 300 words of this page, does it get a complete, confident answer to the most common question in my category?
The brands that invest in this infrastructure in 2026 will not just capture more traffic. They will capture better traffic — pre-qualified, high-intent buyers who have already been recommended by an AI they trust.
The window of competitive advantage is open. But it may not stay open forever.
Key data pint: AI-referred traffic converts 42% better than non-AI traffic (Adobe Digital Insights, April 2026)
Key action: Implement llms.txt, audit passage extractability, build a citation monitor across ChatGPT / Perplexity / Gemini / Claude / AI Overviews. |
To conclude,
Key horizon: By 2027, Gartner projects that 15% of day-to-day business decisions will be made autonomously by AI agents. Your product data needs to be ready for machine consumption, not just human browsing. |
Looking for a comprehensive strategy for your ecommerce business?
Get in touch with me here.
This article is also available in Spanish and German:
Comments