Do ChatGPT, Perplexity and Google use the same ranking signals?

No, and they do not even use the same index. ChatGPT search leans on a Bing-backed index and its own crawler, Perplexity runs its own crawler and index, and Google AI Overviews draw on pages that already rank in Google. That is why a store can appear in one and not the others. They do converge on what they quote: pages that are reachable by crawlers, answer the question in the opening lines, and contain facts that can be checked, such as prices, stock, ratings and policies. Fix those shared factors first, then handle engine-specific issues such as Bing indexing for ChatGPT.

Why does Google AI Overview cite stores that rank on page one only?

Because the overview is generated from pages Google has already retrieved for the query and closely related queries, so the candidate set is essentially the organic results. If your page is not ranking for the question, it is not in the pool the model reads. For a Shopify store this means classic SEO still gates AI visibility on Google: a page per question, a clear title, answer first, internal links from related products, and time. Watch Search Console for impressions on question-style queries; when they rise, overview citations tend to follow. ChatGPT and Perplexity are less tied to Google rank, which is why some stores are cited there first.

Which crawlers should a Shopify store allow in robots.txt?

At minimum Googlebot, Bingbot, OAI-SearchBot (OpenAI's search crawler), PerplexityBot and ChatGPT-User, which fetches pages when a user asks ChatGPT to open a link. Shopify's default robots.txt does not block these, so the usual problem is a merchant who edited robots.txt.liquid to block all AI bots after reading about training crawlers. If you want to opt out of model training but stay visible in answers, block GPTBot specifically and leave OAI-SearchBot allowed; Google's Google-Extended token controls Gemini training without affecting search. Check the live file at yourstore.com/robots.txt after any theme change.

Does a community page really help with AI recommendations, or is it just engagement?

It helps because of what it produces: pages. A customer asking whether a jacket fits over a hoodie, with three replies and a photo, is a page on your domain that answers exactly the kind of question people type into ChatGPT and Perplexity. Multiply that by a few hundred threads and you have a long tail of specific, honest answers that no product description covers. The condition is that those pages are real URLs on your domain, in your sitemap, with structured data, and pinged to Bing. Yourmunity does that by default; a Facebook group or Discord server produces nothing a crawler can read.

How can I tell which engine sent a shopper to my store?

ChatGPT appends utm_source=chatgpt.com to the links it shows and arrives with the referrer chatgpt.com, Perplexity arrives as perplexity.ai, and both appear as separate sources in Shopify Analytics (Sessions by referrer) and Google Analytics 4. Google AI Overview clicks are counted as ordinary Google organic traffic, so there is no separate line; use Search Console impressions on question-style queries as a proxy. Record the landing pages for each source monthly. The pages that receive AI referrals are the ones the engines trust, and they are the template for what to publish next.

How AI search picks which stores to recommend (ChatGPT, Perplexity)

<p>Three assistants now stand between shoppers and your Shopify store: ChatGPT, Perplexity and Google's AI Overviews. They do not share one index or one ranking, which is why a store can be named by Google and ignored by ChatGPT. The direct answer: each engine retrieves a short list of pages from its own index, then has a language model write an answer and cite the pages that supported it, so a store is recommended when its pages are in that index, answer the question plainly and look trustworthy against reviews and policies. This article explains the mechanics of each engine as of October 2026 and the four things they all reward, so you can stop guessing and fix the parts you control in Shopify.</p>

Diagram of how ChatGPT, Perplexity and Google AI Overviews choose stores to recommend

<h2>Quick answer</h2> <p>AI search recommends stores by retrieving pages from an index (Bing-backed for ChatGPT, Perplexity's own crawler, Google's index for AI Overviews), reading them, and citing the ones that answer the question directly with verifiable facts such as price, availability, reviews and policies. To be recommended, a Shopify store needs public, accurate product pages, presence in Bing and Google, and specific question-answering pages on its own domain, such as blog articles and customer community threads.</p>

<h2>How each engine picks</h2>

<h3>1. ChatGPT (search and shopping)</h3> <p>ChatGPT search queries a web index that leans on Bing and reads pages with its own crawler (OAI-SearchBot). For purchase questions it adds merchant feeds submitted through OpenAI's merchant program, which Shopify can supply through its ChatGPT sales channel. It cites a handful of sources, typically pages that state an answer in the first paragraph. If Bing has not indexed your page, ChatGPT cannot cite it, no matter how well it ranks on Google.</p>

<h3>2. Perplexity</h3> <p>Perplexity crawls with PerplexityBot and maintains its own index, then cites many sources per answer, often five to ten. It favors pages with clear structure (numbered steps, tables, FAQs) and recent dates. It shows product cards for shopping intents, drawn from pages and feeds it can read.</p>

<h3>3. Google AI Overviews and AI Mode</h3> <p>Google builds overviews from pages that already rank for the query and related queries, using the regular Googlebot index. The practical consequence is that AI Overview citations follow organic rankings: a page that is nowhere in the top results is rarely cited. Search Console shows these impressions mixed with normal search, so rising impressions on question-style queries is the signal to watch.</p>

<h2>What all three reward</h2> <ul> <li><strong>Being in the index.</strong> Sitemap submitted to Google Search Console and Bing Webmaster Tools, IndexNow for fast pings, no storefront password, no crawler blocks in robots.txt.</li> <li><strong>A direct answer near the top.</strong> First two sentences answer the question; details come after.</li> <li><strong>Verifiable facts.</strong> Price and stock that match the feed, review ratings in structured data (Judge.me, Loox, Yotpo), public shipping and return policies.</li> <li><strong>Specific pages, many of them.</strong> One page per real question: a buying guide, a sizing article, a customer thread about using the product in a particular way. Community pages on your domain (Yourmunity, Forum Innovate, Shopicial) produce these at scale because customers write the questions.</li> </ul>

<table> <thead><tr><th>Engine</th><th>Index it reads</th><th>Sources per answer</th><th>First thing to fix</th></tr></thead> <tbody> <tr><td>ChatGPT</td><td>Bing-backed plus OAI-SearchBot, merchant feeds</td><td>Few</td><td>Bing Webmaster Tools and IndexNow</td></tr> <tr><td>Perplexity</td><td>Own crawler (PerplexityBot)</td><td>Many</td><td>Structured pages: steps, tables, FAQ</td></tr> <tr><td>Google AI Overviews</td><td>Google index, top-ranking pages</td><td>Several</td><td>Rank for question queries first</td></tr> </tbody> </table>

<h2>What this means for a Shopify store</h2> <p>Shopify handles the technical floor: sitemap at yourstore.com/sitemap.xml, canonical tags, Product structured data in modern themes. What it does not do is write the pages that answer questions. A store with 50 products and no blog gives all three engines nothing to quote except product descriptions, and those compete with every other store selling similar items. The stores that get recommended add a layer of question pages: blog articles (by hand, or with autoBlogger, BlogPilot AI, BlogSEO AI or Yourmunity Pro) and customer-written threads on a community page. On GiftX, a store we run, that layer reached about 1,400 Google impressions a day and about 150 clicks a week after six months, and AI citations followed the Google visibility. For how fast it starts, Velpaper, a wallpaper store we run, launched its journal on October 6, 2026 with the same method and had its first clicks within two weeks: 85 clicks and about 7,000 impressions in the week of October 11 to 17, 42 on the best day. Results depend on your niche.</p>

<h2>How to choose what to do first</h2> <p><strong>If you are new to all of this,</strong> start with indexing: Google Search Console, Bing Webmaster Tools with import from Search Console, sitemap submitted to both, IndexNow enabled. It is free and takes an afternoon.</p> <p><strong>If you rank on Google but not in AI answers,</strong> your pages probably bury the answer. Rewrite the first paragraph of your ten most visited pages so the answer comes first, and add a short FAQ.</p> <p><strong>If you have few pages,</strong> pick a publishing routine you can keep: one article a week and a community page where customers add questions. Count pages monthly.</p> <p><strong>If you sell in ChatGPT's shopping flow specifically,</strong> connect Shopify's ChatGPT channel and keep feed data identical to page data; everything else is the same as above.</p>

<h2>Try Yourmunity</h2> <p>Yourmunity is a Shopify app that adds a customer community page on your own domain where every post is an indexed page, pings Bing through IndexNow, and on Pro publishes a weekly SEO article about your products in your store blog. Free, Starter $9, Pro $19 a month, 7-day trial on paid plans. <a href="https://apps.shopify.com/yourmunity">Install Yourmunity from the Shopify App Store</a>.</p>