Free AI visibility checklist for ecommerce brands

When a shopper asks an AI assistant for "the best waterproof hiking boots under $200 with free returns in Canada", the assistant has to decide which stores to name and what to say about them. If your shipping times are stale, your return policy is vague or your products are invisible to its crawler, it will recommend someone else. This checklist covers the fixes we look at first for ecommerce brands.

Use it as a self-audit. Work through each section, tick what's done, and note what isn't. Most items are free to fix; a few need a developer or your ecommerce platform's settings.

Why ecommerce brands should care

Shoppers are already using assistants for product research. In Adobe's survey of 5,000 US consumers, 38% said they had used generative AI for online shopping, and the most common use was research (Adobe, 2025 (opens in a new tab)). AI-referred traffic to US retail sites rose 693% year over year in the 2025 holiday season (Adobe (opens in a new tab)). And the platforms are building shopping directly into answers: OpenAI invites merchants to share product feeds for discovery in ChatGPT, and says Shopify and Etsy catalogs are already integrated (ChatGPT for merchants (opens in a new tab)).

The common thread: assistants recommend stores whose facts they can find, trust and verify. Your job is to make those facts easy to find and consistent everywhere.

1. Baseline: what do AI assistants say today?

• Write 25 to 50 shopper prompts: category ("best…"), comparison ("you vs competitor"), policy ("free returns…", "fast shipping to…"), and brand ("is [store] legit?").
• Run them in ChatGPT, Gemini, Claude, Perplexity and Google (AI Overviews / AI Mode), from your customers' country. Run each prompt more than once.
• Record: are you mentioned, recommended, in what position, and with what attributes (price, quality, shipping, returns)?
• List the competitors that appear and the reasons given for recommending them.
• Note every factual error about you. These are your first fixes.
• Note the sources cited (review sites, articles, marketplaces, forums).

2. Crawl access

• robots.txt does not block search and AI retrieval crawlers you want: Googlebot, Bingbot, OAI-SearchBot (ChatGPT search) and PerplexityBot. See OpenAI's crawler docs (opens in a new tab) and Perplexity's (opens in a new tab).
• Your CDN or bot-protection service (for example a "block AI bots" toggle) isn't blocking them by accident. Check its settings, not just robots.txt.
• Decide separately about training crawlers such as GPTBot or Google-Extended. Blocking training doesn't have to mean blocking search.
• Product, category and policy pages return a 200 status, load without logging in and are in your XML sitemap.
• Important content is in the HTML, not only rendered by JavaScript after interaction, inside images, or in PDFs.
• The site is verified in Google Search Console and Bing Webmaster Tools, with sitemaps submitted and no major indexing errors.

3. Shipping and returns: say it plainly

This is the section that most often decides whether an assistant recommends you or a competitor, because shoppers ask about it directly.

• A dedicated, crawlable shipping page with delivery times and costs by region (for example "Free standard shipping in Canada over $75; 2 to 5 business days to BC, Alberta and Ontario").
• A dedicated returns page with the window in days, who pays return shipping, condition rules, refund timing and exceptions, in plain text.
• Policy summaries repeated on product pages and in the cart, consistent with the policy pages.
• A visible "last updated" date on policy pages, and no old versions still live at other URLs.
• Organization-level return policy structured data (MerchantReturnPolicy (opens in a new tab)) and shipping details, or the equivalent settings in Google Merchant Center. Google added more ways to share shipping and returns policies in late 2025 (Google Search Central blog (opens in a new tab)).
• The same policies stated on marketplaces and listing sites where you sell.
• If your policy is genuinely better than competitors' (longer window, free return shipping), say so clearly. That's exactly the reason assistants give for recommendations.

4. Product data

• Product titles that say what the product is (brand, type, key attribute, size or variant), not just a model name.
• Descriptions that answer the questions shoppers ask: materials, dimensions, fit, compatibility, who it's for, care instructions.
• Price, availability and variants visible in the HTML and kept current.
• Product structured data (opens in a new tab) with offers, price, currency, availability, brand, GTIN or MPN where you have them, and aggregate ratings only if they're genuine and on-page.
• A Google Merchant Center feed with no disapprovals, and free listings enabled.
• If you sell in the US and run on Shopify or Etsy, check how your catalog appears in ChatGPT shopping; other merchants can apply to share a feed (OpenAI (opens in a new tab), feed spec (opens in a new tab)).
• Size guides, comparison tables and FAQs on high-value product and category pages.

5. Brand facts and consistency

• An About page stating what you sell, who for, where you ship from and to, and how long you've operated, in plain sentences.
• Organization structured data (opens in a new tab) with your official name, logo, contact and social profiles (sameAs).
• The same name, description and category on Google Business Profile (if you have a location), social profiles, marketplaces and directories.
• Clear contact and customer service details. Assistants are asked "is this store legit?" more than you'd think.
• Optional: an llms.txt (opens in a new tab) file summarizing your store, policies and key pages. Low cost, uncertain benefit.

6. Reviews and third-party sources

• A steady flow of genuine reviews on the platforms engines cite for your category (often Google, Trustpilot, marketplaces or niche review sites). Follow each platform's rules on asking for reviews.
• Responses to negative reviews, especially about shipping, returns or customer service, that state the current facts politely.
• Presence in the "best of" and comparison articles that assistants cite for your top prompts. Contact authors with samples, data or corrections.
• Answers in relevant communities (Reddit, forums) from clearly identified staff, focused on helping rather than selling.
• Any outdated third-party claims about you (old prices, old policies) listed and corrected at the source where possible.

7. Answer-first content

• A buying guide for each major category that answers the questions in your prompt set.
• Honest comparison pages for your most common "[you] vs [competitor]" questions.
• FAQ sections with short, direct answers, on product, category and policy pages.
• Specific numbers and sources where they help (test results, materials data, certifications you actually hold).
• Content reviewed by someone who knows the products. Generic AI-written copy rarely gets quoted.

8. Tracking and reporting

• An "AI assistants" channel group in your analytics that captures referrals from chatgpt.com, perplexity.ai, gemini.google.com, copilot.microsoft.com, claude.ai and similar.
• Conversions and revenue from that channel tracked like any other.
• Your prompt panel re-run on a schedule (we recommend weekly), with mention rate, recommendation rate, share of voice and accuracy tracked over time.
• A simple log of changes you made (policy updates, new pages, review pushes) so you can connect actions to movement.

How to prioritise

If you only have a day, do these five things first:

1. Run 25 prompts across four assistants and write down every error about your store.
2. Check robots.txt and your CDN settings for blocked search and AI crawlers.
3. Rewrite your shipping and returns pages in plain, specific language, add a "last updated" date, and remove old versions.
4. Add or fix Organization, Product and return policy structured data.
5. Claim and complete your profiles on the three review or listing sites engines cite most for your category.

Then revisit the panel in two to four weeks. Retrieval-heavy engines like Perplexity and Google often reflect fixes sooner than answers that lean on older training data.

Common pitfalls

• Blocking every AI bot in a panic, including the ones that power search answers, then wondering why you disappeared.
• Policies that differ between the site, the cart, marketplaces and old blog posts. Assistants pick one, not always the right one.
• Fake or incentivized reviews. Against platform rules, and the kind of signal you don't want amplified.
• Treating one screenshot as truth. Measure rates across repeated runs.
• Chasing tricks. Hidden text and "prompt injection" in page copy are more likely to get you ignored than recommended.

Frequently asked questions

Want this done for you? Our AI visibility service runs the prompt panel weekly and fixes what assistants get wrong. Start with a free audit. For the strategy behind the checklist, read how to get ChatGPT and Perplexity to recommend your brand.