How to rank in ChatGPT Shopping, Perplexity and Google AI Overviews
Fashion brands must implement structured product data, optimise for natural language queries and measure visibility across AI assistants to compete for agentic commerce traffic in 2026.

KEY TAKEAWAYS Summary by the editors
- Google AI Overviews now appear in 85 per cent of US ecommerce searches, up from 24 per cent in April 2026, whilst ChatGPT shopping is live for US users and Perplexity has reached a reported 2 billion dollars annualised GMV.
- AI agents prioritise structured data in JSON-LD format, requiring brands to implement Product, Offer and AggregateRating schema with complete attributes including GTIN, brand, price, availability and imagery.
- Perplexity weights schema markup at 28 per cent and quotability at 18 per cent in its commerce ranking algorithm, whilst Google AI Overviews cite pages ranking in the top 20 organic results 56 per cent of the time.
- Brands should track AI visibility by monitoring citation frequency, inclusion rate and referral behaviour across ChatGPT, Perplexity and Google using platform-specific measurement tools or daily query audits.
Fashion brands that want to appear in AI shopping recommendations must optimise structured product data, implement schema markup and track visibility across conversational search platforms. Google AI Overviews now appear in 85 per cent of US ecommerce searches, up from just 24 per cent in April 2026, whilst ChatGPT shopping launched on 24 March 2026 and is live for ChatGPT users in the US. Perplexity's commerce division is processing what internal sources describe as a 2 billion dollar annualised gross merchandise value run rate as of June 2026. The optimisation principles differ from traditional search engine optimisation because AI agents read structured data rather than visual layouts and generate shortlists of three to five products rather than pages of results.
What structured data do AI shopping agents require?
Product schema refers to structured data markup based on the Schema.org vocabulary that describes product information in a machine-readable format, typically added to a product page's HTML as JSON-LD. The core types are Product (the item: name, reference such as SKU or GTIN, brand, description, attributes) and Offer (the commercial offer: price, currency, availability, delivery and return terms). At minimum, every SKU needs a baseline of core attributes: title, description, brand, GTIN, MPN, category, price, sale price, availability, condition, image URL and product URL.
OpenAI's product feed schema requires nine fields in basic product data: description (factual product description for this item) and brand (product brand as shown on the product page), along with identifier, link, title, image link, price, availability and condition. The citation criteria for shopping AI Overviews differ from informational AI Overviews: product availability, pricing transparency and structured data markup became important additional signals, and sites without proper product schema were largely excluded from shopping AI Overviews regardless of their domain authority.
Which product attributes matter most for fashion?
Fashion products require category-specific attributes beyond the baseline schema. Colour is an optional string attribute, selected colour consistent with the product image, whilst size, material, fit, care instructions and occasion data allow AI agents to match natural language queries to inventory. Fashion brands that tag images with detailed alt text describing colour, pattern, silhouette, styling context and occasion create a dual optimisation layer for both current text-based and emerging visual AI recommendation systems.
Perplexity weights five signals for commerce queries: schema (28 per cent), freshness (22 per cent), external corroboration (18 per cent), in-prose quotability (18 per cent) and entity clarity (14 per cent), with schema and quotability fully under merchant control and covering 46 per cent of the ranker. A March 2026 study from Peec AI found that about 83 per cent of products in ChatGPT carousel positions also appeared in the Google top 40 products, suggesting that optimisation for Google Shopping feeds remains a foundation for AI visibility.
How should brands optimise content for AI agents?
AI agents cannot interpret visual design but can parse structured text and schema. Beautifully designed product pages stimulate human purchase intent, but AI agents cannot see that design; what agents reference is structured data written in JSON-LD, field values returned from API endpoints and attribute information registered in Merchant Center. It is not about adding text for the sake of it, as AI more easily cites text that directly answers a question using concrete data.
Product descriptions should use natural language that mirrors how shoppers search. Conversational queries like "black wool coat under £300, ships by Friday" map cleanly to structured product attributes, making apparel retailers some of the highest-converting AI shopping traffic in 2026. Research from Profound shows that 60 per cent of AI shopping queries begin without a specific brand name, meaning generic discovery queries such as "sustainable work trousers" or "wedding guest dress summer" drive visibility more than branded searches.
What is the role of external signals and reviews?
Visibility on AI assistants does not depend on traditional advertising, but on digital reputation built through independent external sources such as trade press, reviews and forums. AggregateRating (average rating and review count) and individual review objects provide social proof data that AI systems weigh when making recommendations. In November 2025, 37.87 per cent of AI Overview sources ranked in the top 20 organic results, but by September 2026 that share reached 56.38 per cent, indicating that traditional SEO authority still influences AI citation eligibility.
The State of Fashion 2026 report by McKinsey and Business of Fashion describes fashion brands' presence in AI chatbot responses as the new SEO. To ensure their products are visible and favoured by AI models, brands must rethink their digital marketing and ecommerce infrastructure, where semantically rich data and API-accessible content will be critical to success.
How do the three platforms differ in ranking logic?
ChatGPT shopping is a product discovery experience, and product results are organic and unsponsored; as of 2026 the shopping feature shows no advertising and no paid placement, and products appear based on relevance to your request. Walmart introduced an in-ChatGPT app experience on 24 March 2026 that takes users from discovery in ChatGPT into a tailored Walmart environment that supports account linking, loyalty and Walmart payments, demonstrating the custom integration path available to large retailers.
Perplexity Commerce launched in beta in Q4 2025 and moved to general availability in March 2026, embedding shoppable product cards directly inside Perplexity's AI-generated answers. Perplexity Shopping was launched in beta in November 2025 and expanded internationally in Q1 2026, allowing users to search, compare and purchase products directly from the Perplexity interface without visiting the merchant's site. Google AI Overviews differ in that 83 per cent of "best [product]" queries now show AI Overviews, with queries like "best running shoes 2026" or "best wireless earbuds under £100" almost guaranteed to have an AI Overview at the top.
How should fashion brands measure AI visibility?
McKinsey stated that only a limited number of brands currently track AI-search performance systematically, creating a major opportunity for early movers, and those brands are beginning to monitor AI citation frequency, inclusion within recommendation consistency across platforms, AI-generated summaries and how brand sentiment is represented inside LLM responses. Monitoring brand visibility in 2026 should separate five outcomes: inclusion, accuracy, citation, sentiment and referred behaviour.
Practical measurement requires asking AI platforms the same queries shoppers use. Track how ChatGPT, Gemini, Claude and Perplexity answer 20 to 40 of your shoppers' real questions, ideally daily, and read the trend weekly. Set up custom GA4 segments filtering for AI user agents (ChatGPT-User, PerplexityBot) and compare conversion rates between AI-referred and organic sessions. Launchmetrics rolled out its new metric AI Visibility (AIV) alongside New York Fashion Week to measure a brand's impact, and brands will be able to track how they are featured within LLM responses on platforms like Google Gemini and ChatGPT.
| Platform | Key ranking signals | Measurement approach |
|---|---|---|
| ChatGPT Shopping | Google Shopping feed overlap, product schema, server-side rendering | Track product mentions in conversational queries; monitor ChatGPT-User referrals in GA4 |
| Perplexity Commerce | Schema markup 28%, freshness 22%, external corroboration 18%, quotability 18% | Submit product feed via merchant programme; track PerplexityBot referrals and citation rate |
| Google AI Overviews | Top 20 organic ranking, Product schema, pricing transparency, reviews | Monitor AI Overview appearance rate on target queries; track citation sources in Search Console |
Similarweb measured AI referrals up 278 per cent in fashion and 312.5 per cent in beauty year over year (June 2025 to May 2026), and Adobe found AI-referred visitors to US retail sites converting 54 per cent better than other visitors. The optimisation effort is incremental to, not a replacement for, traditional search and paid channels, but the growth rate and conversion quality justify dedicated resource allocation for most fashion brands operating in English-language markets where these platforms have launched.
Frequently asked questions
Do fashion brands need different structured data for each AI platform?
No. Schema.org Product and Offer markup in JSON-LD format serves all three platforms. ChatGPT also accepts direct product feed submission, Perplexity offers a merchant feed programme launched in April 2026, and Google reads both schema and Merchant Center data. Implementing standards-compliant schema once makes products eligible across platforms.
Can brands pay to rank higher in ChatGPT or Perplexity shopping results?
No for ChatGPT, which shows organic and unsponsored results as of 2026. Perplexity operates a zero-fee merchant programme with no cost to appear or sell. Google AI Overviews may show ads alongside the AI-generated summary, but the citations within the overview itself remain algorithmic. Paid placement is not available in the core recommendation logic of any platform.
How quickly can brands expect to see results from AI optimisation?
Schema implementation takes effect within days once crawled, but citation eligibility depends on broader authority signals. Brands report measurable AI referral traffic within six months of implementing complete product schema, optimising for natural language queries and building external validation through press and reviews. Most fashion brands see AI traffic as 8 to 15 per cent incremental acquisition within six months.
