When ChatGPT recommends a product, it’s increasingly not reading your product page at all. It’s reading your product feed — a structured file most SEO teams have never opened.
Google’s Merchant Center feed specification lays out the exact attribute requirements referenced above.
Quick answer: Feed-sourced retrievals grew from 4.3% to roughly 20% of ChatGPT shopping retrievals in just six weeks in 2026, and feed-based offers show up as the top product 99.9% of the time when a feed exists. Around 88% of ChatGPT’s product offers still originate from product detail pages rather than feeds today, but that share is shrinking fast. Feed data gives AI shopping systems 100% consistent brand, image, and merchant fields, versus 0% consistency for page-scraped offers — which is exactly why feed completeness is becoming the new ranking factor for AI shopping.
Product feed SEO: feeds are winning the discovery layer
Research into where ChatGPT sources its product recommendations found that 83% of its picks matched Google’s top 40 organic Shopping results — meaning AI shopping isn’t inventing a new ranking system from scratch, it’s largely inheriting an existing one built on product feed data. When a feed-based offer exists for a product, it appears as the top result 99.9% of the time, because feed data arrives in clean, structured fields that AI systems can trust without having to interpret and reconcile inconsistent page content.
Feeds vs. pages, by the numbers
| Metric | Value |
|---|---|
| Feed-sourced ChatGPT retrievals, six-week growth | 4.3% → ~20% |
| ChatGPT product offers still from product pages | ~88% |
| Feed-based offers appearing as top product | 99.9% |
| Field consistency: feed data vs. page-scraped data | 100% vs. 0% |
| Strong matches originating from top-10 Shopping results | 60% |
Read those two headline numbers together and the direction is clear: product pages still supply the majority of what ChatGPT shows today, but feeds are capturing share at a pace that suggests they’ll be the primary discovery layer for AI shopping well before this decade is out.
Why AI traffic is worth building this for
This isn’t a hypothetical channel. AI-sourced visits to US retail sites grew 393% year-over-year in Q1 2026, and AI-linked sales reached an estimated $262 billion during the holiday shopping season. Retailers who treat their Google Merchant Center feed as a secondary, set-and-forget data source are leaving a fast-growing, high-intent channel almost entirely to chance.
The feed attributes that actually matter
Core requirements haven’t changed: a valid GTIN, an accurate title, current price and availability, a clean image, correct brand, and the right product category. What’s new in 2026 is a set of optional conversational attributes built specifically for AI shopping assistants:
- Question-and-answer pairs — pre-formatted answers to common product questions, ready for an AI assistant to surface directly.
- Related product relationships — explicit links between complementary or alternative products.
- Document links — manuals, spec sheets, and other supporting documentation an AI shopping agent can reference for detail.
- Item group title and variant options — clean structuring for products that come in multiple sizes, colors, or configurations.
- Popularity rank scores — a structured signal of relative demand across your catalog.
Product pages still matter — just not the way you think
None of this makes your product detail pages obsolete. Pages remain essential for building credibility, hosting genuine reviews, and closing the sale once an AI system has pointed a shopper your way. The shift is in what determines initial visibility and ranking: feed completeness now does that job, while the page itself does the convincing once a shopper actually arrives.
How to prepare your feed for AI shopping
- Audit your Google Merchant Center feed for completeness on the core fields first — GTIN, brand, price, availability, and category — since incomplete core data disqualifies you before the conversational attributes even get considered.
- Add the new conversational attributes where your platform supports them, starting with Q&A pairs and item group/variant data for your best-selling products.
- Keep price and availability in real time. Feed data that lags your actual inventory undermines the trust advantage feeds are supposed to provide.
- Don’t neglect your product pages. They still carry the majority of ChatGPT’s product offers today, and they remain where reviews and credibility signals live.
- Re-check your feed monthly given how fast feed-sourced retrieval share is moving — a stale audit from early 2026 is already out of date.
Frequently asked questions
Both, but the mix is shifting. Around 88% of ChatGPT’s current product offers still come from product detail pages, but feed-sourced retrievals grew from 4.3% to about 20% in just six weeks in 2026 — and when a feed-based offer exists, it appears as the top result 99.9% of the time.
The core fields come first: a valid GTIN, accurate title, current price and availability, a clean image, correct brand, and category. Only after those are complete do the newer conversational attributes — Q&A pairs, variant data, popularity rank — add meaningful additional visibility.
Yes. Feeds are increasingly winning the discovery and ranking layer, but product pages still supply most of ChatGPT’s current offers and remain where reviews, detailed specs, and trust signals live once a shopper clicks through.
Strongly. AI-referred traffic to US retail sites converted 42% better than non-AI traffic in early 2026, and AI-sourced visits grew 393% year-over-year in Q1 alone, with AI-linked sales reaching an estimated $262 billion over the 2025 holiday season.
New, optional Google Merchant Center fields built for AI shopping: question-and-answer pairs, related product relationships, document links like manuals or spec sheets, item group and variant data, and popularity rank scores. None are required, but each gives an AI shopping agent more structured context to work with.
Get your product feed ready for AI shopping
Most retailers have never audited their product feed the way they audit their website. Get a free website audit and we’ll show you exactly which feed attributes are missing, broken, or holding your products back from AI shopping visibility. Ready to capture this channel before your competitors do? Talk to our team and we’ll take it from there.
Related Reading
- Shopify SEO Guide (2026)
- Agentic Commerce: How AI Shopping Agents Are Starting to Buy in 2026
- Best Shopify Apps in 2026
Riya Bhardwaj
Lead Marketing Strategist
Leading content and growth initiatives with a focus on search visibility, audience engagement, and measurable business outcomes. Specialised in SEO, Generative Engine Optimization (GEO), AI search optimisation, and performance-driven content marketing. Passionate about transforming market insights into scalable content strategies that strengthen brand authority and drive sustainable digital growth.