Okay, I need to talk about something that's going to annoy a lot of content-first SEO teams. And I say this as someone who genuinely loves content-first SEO — I've evangelised it, I've bored people at parties with it. (The parties were already bad. I didn't help.)

But the data is in, and it's saying something uncomfortable: for any query where money changes hands, the single most important AEO/GEO signal isn't your content quality, your topical authority, or your carefully constructed E-E-A-T signals. It's your structured data. Your machine-readable product feeds. The unglamorous spreadsheet work that nobody puts in their LinkedIn bio.

I know. I didn't want it to be spreadsheets either. And yet, here we are.

Let me give you the evidence before the argument, because the evidence is doing most of the heavy lifting here.

An analysis of over 4,000 keywords and nearly 40,000 product grids across nine months found that product grid placements on Google grew 82% in that period. Ninety-six percent of all SERPs in the dataset now show at least one product grid for commercial queries. The traditional organic result — the one most of us have spent our careers optimising for — has been physically pushed below the fold.

That alone should make you sit up. But here's where it gets properly interesting for anyone thinking about answer engine optimisation.

Research from Peec AI decoded a hidden field in ChatGPT's source code — a base64-encoded parameter — and discovered that ChatGPT's product carousels are being populated directly from Google Shopping results. Not from organic rankings. Not from Bing. Not from some independent AI-curated product index. From Google Shopping's structured data layer.

The match rate: 83% of ChatGPT's carousel products aligned with Google's top 40 organic shopping results. For Bing, it was 11%. There were 70 products in the entire 43,000-product dataset that appeared exclusively in Bing. Seventy. That's not a secondary data source — that's a rounding error with feelings.

And there's a clear positional bias: 60% of strong matches came from the top 10 Google Shopping results, nearly 84% from the top 20. Rank higher in Google's structured shopping index, appear earlier in ChatGPT's carousel. This pattern held across all ten product verticals tested and across both branded and non-branded prompts.

So: structured product data determines your Google Shopping ranking, which determines your product grid visibility, which now determines your visibility in ChatGPT's product recommendations. One input. Three surfaces. That's not a nice-to-have optimisation. That's the new AEO infrastructure.

Here's why this matters beyond shopping. The ChatGPT-Google Shopping connection isn't just a quirky implementation detail — it's revealing how large language models solve the problem of recommending specific things.

When an AI needs to answer a subjective question — "explain quantum entanglement" or "what caused the fall of Rome" — it can synthesise training data and generate a plausible response. Messy, maybe. Debatable, often. But functional.

When it needs to answer a commercial question — "what's the best running shoe for flat feet under £120" — synthesis isn't enough. It needs current prices, real availability, actual product specifications, merchant ratings. It needs structured, machine-readable, trust-verified data. And right now, the richest source of that data is Google's Merchant Center ecosystem.

This is the AEO insight hiding in the shopping data: AI systems don't generate product answers from content. They retrieve them from structured data indices. The model that best describes how ChatGPT handles commercial queries isn't "read the web and summarise." It's "query a structured database and display results."

And if you think this retrieval pattern will stay confined to product recommendations, I have some very reasonably priced beachfront property to sell you. Every category where AI needs to recommend something specific — local businesses, professional services, software tools, travel options — will eventually follow the same structural logic. The system that can serve clean, machine-readable, verified data to the AI wins the citation.

If you're used to thinking about AEO in terms of "create comprehensive content that answers questions well," this is going to feel like a gear change. For commercial and recommendation-type queries, the signals that determine AI visibility look much more like marketplace optimisation than content strategy.

Feed quality over content quality. Your Merchant Center data — product titles, GTINs, structured attributes, error rates — matters more than your on-page copy for grid and carousel placement. Google doesn't need to crawl your HTML to populate a product grid. It needs a clean, correctly formatted data feed. The AI doesn't read your product page. It reads your schema.

Visual assets over backlinks. A high-quality product image on a white background generates more engagement in grids than a dozen referring domains. In the structured data environment, image quality, price competitiveness, and merchant ratings are the visible competitive levers. Domain authority is invisible. (I realise I just told an entire profession that their primary metric is invisible. I'm sorry. I'm also right.)

Price competitiveness over domain authority. When users — and AI systems — can compare prices at a glance, a fifteen-year-old domain carries no weight. The lowest competitive price with acceptable shipping terms wins. This commoditises visibility in a way traditional SEO never did.

Category pages over product pages. The analysis found 97% of product grid placements went to category and listing pages, not individual product detail pages. Google's structured shopping index overwhelmingly favours filterable catalogue views. For AI retrieval, structured breadth beats individual depth.

Perhaps the most dangerous dimension of this shift is that most teams literally cannot see it in their own analytics.

Google Search Console reports on traditional organic results. Merchant Center provides its own product-level data. But there is no unified view that answers the fundamental AEO question: what share of our total AI-surface visibility comes from structured data feeds versus traditional organic content?

Without third-party tooling, the split is invisible. Discount Computer Depot can look at Search Console and see market-leading performance while completely missing that those rankings are increasingly irrelevant for the queries that generate revenue. The dashboard says you're winning the AEO game. The actual SERP — and ChatGPT's carousel — say otherwise.

This creates a compounding problem. Because teams can't see the structured data gap, they keep allocating resources to content optimisation. Pages get refined, articles get updated, link campaigns continue — all valuable work, but increasingly aimed at a surface that product grids and AI carousels are pushing further from view.

The AEO/GEO conversation has been dominated by content strategy — write better answers, build topical authority, earn citations through comprehensive coverage. And for informational queries, that still holds. Nobody's arguing otherwise.

But for any query where AI needs to recommend something specific — a product, a service, a place, a tool — the dominant signal isn't content. It's structured data. Google's product grids grew 82% in nine months. They appear on 96% of commercial SERPs. And ChatGPT is pulling its product recommendations directly from Google's structured shopping index, not from your organic rankings.

The brands that invest in structured data quality now aren't just winning on Google. They're building a compounding advantage across every AI surface that retrieves from the same index. The brands that don't will keep optimising content for a surface that's being quietly relocated below the fold — which is a bit like perfecting your window display after the high street moved.

Your Merchant Center feed is your AEO strategy. Time to start treating it like one.

Jo Lambadjieva is an entrepreneur and AI expert in the e-commerce industry. She is the founder and CEO of Amazing Wave, an agency specializing in AI-driven solutions for e-commerce businesses. With over 13 years of experience in digital marketing, agency work, and e-commerce, Joanna has established herself as a thought leader in integrating AI technologies for business growth.

😍It was great, all I needed

🙂Good, not great

🙄It sucked