top of page

Ranking #1 on Google Doesn't Mean AI Recommends You

21 hours ago
5 min read

Short answer: No. A brand can own the top organic spot on Google and still be missing, misdescribed or outranked by competitors in ChatGPT, Perplexity, Gemini and Claude. When AI engines can't read a brand's own website clearly, they fill the gap with third-party sources, and those sources are often incomplete, outdated or financially motivated.

Over the past year, Bantage has run AI visibility studies for major brands in travel, financial services and retail. We ran the questions real customers ask (which card to get, which airline to fly, where to book, what to buy) across the major AI search engines, then traced where each answer came from. We've researched hundreds of brands, thousands of prompts or queries in fifteen different languages.

Three patterns showed up again and again, and none of them show up in a traditional SEO report.


Finding 1: If AI can't read your site, it will get its information elsewhere


AI engines answer from what they can retrieve and understand. When a brand's own site doesn't state the key facts clearly (what it offers, who it's for, what it costs, how it compares), the engine doesn't stop. It looks elsewhere.

That "elsewhere" is review sites, comparison pages, forums, old press coverage and aggregator listings. In our studies, these third-party sources were regularly:

  • Incomplete: missing current products, features or markets.

  • Outdated: citing old fees, discontinued offers or retired product names.

  • Inaccurate: getting basic facts about the brand wrong. For example, we found that an airline's baggage fees were quoted from a 2022 travel agency article that was out-of-date.

The brand had the right answer on its site. It just wasn't published in a form the engine could find, parse and trust, so a stranger's version became the official one.


Finding 2: Obscure review sites have outsized influence


The sources AI engines lean on are not always the big, familiar names. Again and again, we found relatively obscure, niche review sites shaping the answer, sometimes cited more often than the brand itself.

Many of these sites fund themselves through affiliate revenue. Some are sub-affiliates: they earn commissions through a network or another partner rather than a direct relationship with the brand. That creates three problems:

  • They're off the radar. Marketing and affiliate teams track their direct partners. Sub-affiliates often don't appear in any report the brand sees.

  • Their incentives are unclear. Rankings and recommendations can follow commission rates rather than product relevance, quality or fit.

  • Nobody keeps them up-to-date. With no direct relationship, there's no one to correct a dated fee table or a wrong product description.

The result is that a small site with questionable, dated or biased information can quietly become one of the most influential voices describing your brand to AI search users. We saw a marine lifestyle site having outsized citation weight in how financial products were positioned and described.


Finding 3: Content built for humans can end up "in the drawer"

Many brand websites are designed for people clicking around: tabs, accordions, carousels, "load more" buttons and embedded widgets. That works well for a visitor. But content that only appears after a click, or that loads from another source, can be invisible to some or all AI crawlers.

We call this content "in the drawer." It exists, a human can open the drawer, but many AI engines never do. Common culprits:

  • Tabbed sections where only the first tab ships with the page.

  • Embedded HTML widgets that load from a separate domain inside an iframe.

  • Content rendered by JavaScript after the page loads.

  • Key facts inside images or PDFs instead of page text.

We found this on our own site. Bantage's FAQ had 16 questions organized into category tabs, but only the first tab's 6 questions were in the page's HTML. Our answers about AI visibility, the topic we care most about, were in the drawer. Our homepage's product overview had the same problem: it was built as an embedded widget, so none of its text was on the page crawlers read. Both took an afternoon to fix once we knew to look.

If it happens to a company that measures AI visibility for a living, it's worth checking your own site.


Why your Google ranking doesn't carry over


Google's organic rankings reward authority, links and relevance built up over years. AI engines work differently. They retrieve passages, judge which sources answer the question most directly, and assemble a response. A strong domain helps, but it doesn't guarantee that your facts are the ones that get retrieved and quoted.

And every engine is different. In our studies it was common for a brand to be cited consistently by one engine and absent from another for the same question. A single "AI score" hides that.


What to do about it


  1. Check what crawlers actually see. View your key pages' raw HTML, not the rendered page. If important text only appears after a click or loads in an embed, move it into the page.

  2. State the basics in plain text. What you are, who you serve, what you offer and how you're different, written as clear sentences on your homepage, product pages and FAQ.

  3. Add structured data. Organization, product and FAQ schema help engines connect your facts to your brand.

  4. Find the sources engines cite about you. Run your customers' real questions across ChatGPT, Perplexity, Gemini and Claude, and list every site that shows up.

  5. Map your affiliate ecosystem, sub-affiliates included. Know who is describing your brand for a commission, and ask your networks for visibility below the first tier.

  6. Correct the record. Update the third-party pages you can influence, and publish authoritative answers for the ones you can't.

  7. Measure by engine, over time. AI answers shift as models and sources change. Track visibility per engine, not as one number.

Frequently asked questions

Does ranking #1 on Google mean ChatGPT will recommend my brand? No. AI engines retrieve and weigh sources differently from Google's organic rankings. Top-ranked brands can still be missing or misdescribed in AI answers.

Why do AI engines cite third-party sites instead of my website? Usually because the facts aren't stated clearly on your own site, or they're hidden in tabs, embeds or images that crawlers don't read. Or the site is blocked inadvertently.

What is a sub-affiliate, and why does it matter for AI search? A sub-affiliate earns commissions through a network or partner rather than directly from the brand. These sites often go untracked by marketing teams, yet AI engines may cite them heavily.

How do I check whether my content is "in the drawer"? View your page's raw HTML and search for key sentences. If they're missing, AI crawlers likely can't see them either.


See where your brand stands


Bantage Brand Presence tracks how ChatGPT, Perplexity, Gemini and Claude describe your brand, which sources they rely on, and where competitors are cited instead. Request a free AI visibility analysis or learn more about Brand Presence.

Noah Tratt is the founder of Bantage. He spent 17 years at Expedia leading SEO and SEM and built Expedia's advertising business.

 
 
 

Comments


bottom of page