How to audit your fashion brand's visibility in ChatGPT, Gemini and Perplexity
A repeatable audit shows whether AI assistants mention your brand, what they say about it and which pages they cite. Here is a practical method, from prompt sets to crawler checks.
KEY TAKEAWAYS Summary by the editors
- An AI visibility audit tests a fixed set of shopper and buyer prompts in several assistants and records whether the brand is mentioned, how it is described and which sources are cited.
- OpenAI states that sites which block OAI-SearchBot will not be shown in ChatGPT search answers, while blocking GPTBot only signals that content should not be used for model training.
- Google says there are no extra technical requirements or special schema.org markup needed to appear in AI Overviews or AI Mode; a page must be indexed and eligible for a snippet.
- Assistant answers vary between runs, accounts and locations, so an audit should repeat prompts several times and report shares of answers rather than single screenshots.
- The audit is only useful if every finding is mapped to an owner and a fix, such as correcting product data, publishing missing facts or earning coverage on sources the assistants already cite.
To audit your fashion brand's visibility in ChatGPT, Gemini and Perplexity, run a fixed set of real customer prompts in each assistant, record whether your brand appears, how it is described and which pages are cited, then check that the crawlers behind those assistants can actually reach your site. Repeat the exercise on a schedule, because answers change from week to week.
The audit is not a ranking report. Generative assistants compose an answer, often by running several searches in the background, and then cite a handful of sources. What you are measuring is your share of answers, the accuracy of what is said about you and the sources that shape it.
Why should a fashion brand audit its AI visibility now?
Shoppers are increasingly starting product research inside assistants rather than on a results page. Adobe reported that traffic from generative AI tools to US retail sites rose 693.4% year on year between 1 November and 31 December 2025, while noting that the base of users remains modest. McKinsey and The Business of Fashion, in The State of Fashion 2026, describe AI chatbot responses as "the new SEO" and say brands will need semantically rich, API-accessible content to stay visible to AI models.
If an assistant describes your brand with an outdated price position, a discontinued line or a competitor's product, you want to know before customers do.
Which prompts should the audit test?
Build a prompt set of 30 to 60 questions that mirror how your customers actually ask. Group them by intent so that results can be compared over time:
- Category discovery: "best merino base layers for hiking", "sustainable denim brands for women".
- Brand questions: "is [brand] good quality?", "where is [brand] made?", "what is [brand]'s return policy?".
- Comparisons: "[brand] vs [competitor] for tailored trousers".
- Product specifics: sizing, fit, materials, care and price for named hero products.
- Local and channel questions: "where can I buy [brand] in Zurich?", "which stores stock [brand]?".
Use your search console queries, customer service logs and on-site search terms as raw material. Write the prompts in each language and market you sell in, because assistants draw on different sources by language.
How do you run the prompts so the results are reliable?
Assistant answers are not deterministic. The same prompt can produce different brands on different days, in different accounts or with personalisation features such as memory switched on. Reduce the noise with a simple protocol:
- Use clean sessions: logged-out or fresh test accounts with memory and personalisation disabled where the product allows it.
- Run each prompt at least three times per assistant and record every run.
- Fix the location and language settings for each market.
- Record the full answer text, the brands named in order, any product cards shown and every cited URL.
- Store results in a spreadsheet or database with the date, assistant, mode (for example search on or off) and prompt ID.
From this you can calculate a mention rate (share of runs in which the brand appears), a first-mention rate, a citation rate (share of runs in which one of your own pages is cited) and an accuracy rate based on a manual check of the claims made.
What should you score in each answer?
| Metric | How to measure | What it tells you |
|---|---|---|
| Mention rate | Share of runs that name the brand | Whether you are part of the consideration set |
| Position | Order of first mention among brands | Relative prominence against competitors |
| Owned citations | Share of runs citing your domain | Whether your own pages are trusted as sources |
| Third-party citations | Domains cited when you are mentioned | Which media, retailers or forums shape the answer |
| Accuracy | Manual check of prices, materials, policies, stockists | Risk of misinformation reaching customers |
| Sentiment and framing | How the brand is characterised | Positioning drift, for example premium described as budget |
The third-party citation list is often the most actionable output. If assistants consistently cite a particular retailer, review site or trade publication when discussing your category, those are the sources whose information about you must be correct and current.
How do you check that AI crawlers can reach your site?
Visibility starts with access. Each provider documents separate crawlers with separate purposes, and a single blanket block in robots.txt or at the content delivery network can remove a brand from answers without anyone noticing.
- OpenAI: OAI-SearchBot surfaces sites in ChatGPT search; OpenAI states that sites opted out of it will not be shown in ChatGPT search answers. GPTBot collects content that may be used for training, and OpenAI says each setting is independent.
- Perplexity: PerplexityBot is designed to surface and link websites in Perplexity results; Perplexity-User fetches pages when a user asks a question.
- Google: AI Overviews and AI Mode rely on normal Googlebot crawling and indexing. Google-Extended controls use for training and grounding in other Google systems, not inclusion in Search.
Check robots.txt, firewall and bot-management rules, and server logs for these user agents. Confirm that product pages render key facts (price, material, sizes, availability) in HTML text rather than only in images or client-side scripts.
What should you do with the audit findings?
Turn each gap into a task with an owner. Inaccurate facts usually trace back to product data, outdated pages or third-party listings, so the fixes sit with e-commerce, PIM and PR teams as much as with marketing. Typical actions include publishing clear pages for policies, sizing and materials, keeping structured data consistent with visible content, updating retailer and marketplace listings, and earning coverage on the sources the assistants already cite.
Be realistic about what an audit can prove. It measures a sample of answers at a point in time; it does not reveal how many people asked those questions. Pair it with analytics data on referrals from assistants and with Google Search Console, where Google says AI Overview and AI Mode traffic is included in the overall Web search figures.
How often should the audit be repeated?
A quarterly full audit with a monthly check of 10 to 15 priority prompts is a workable rhythm for most brands. Re-run the relevant prompts after major launches, price changes, site migrations or changes to robots.txt. Keep the prompt set stable so trends are comparable, and add new prompts in a separate group rather than replacing old ones.
Frequently asked questions
How do I check if ChatGPT mentions my brand?
Write a fixed set of realistic customer questions, run each several times in a clean ChatGPT session with search enabled and record whether your brand is named and which pages are cited. A single test is not enough because answers vary between runs.
Does blocking GPTBot remove my site from ChatGPT search?
No. OpenAI documents GPTBot as the crawler for training data and OAI-SearchBot as the crawler for search. OpenAI states that sites opted out of OAI-SearchBot will not be shown in ChatGPT search answers, and that each setting is independent.
Do I need special markup to appear in Google AI Overviews?
Google says there are no additional technical requirements and no special schema.org structured data needed for AI Overviews or AI Mode. Pages must be indexed and eligible to show with a snippet, and structured data should match visible content.
What tools can I use to track AI visibility?
You can start with a spreadsheet, a stable prompt set and manual runs, then add analytics for AI referrals and Search Console data. Commercial monitoring tools automate prompt runs, but the method and the prompt set matter more than the tool.
One edition every weekday morning. Read in five minutes. Free for industry professionals.
