500 AI Searches Later: Here's Which Brands and Media Outlets LLMs Actually Cite

Try searching for "best travel credit cards" in Google versus ChatGPT and you'll see two completely different worlds.

The Google Reality

When I search Google, the first two results are ads. Below that? Ten blue links staring back at me. Ten potential tabs to open, ten sources to compare, ten rabbit holes to fall down. And here's the kicker: each of those tabs will have different answers and rankings, forcing me to spend even more time cross-checking recommendations for each card.

And let's be honest, when was the last time you clicked to page 2 on Google? For most of us, the answer is never. Those first ten results essentially make up the internet. This friction is exactly why more people are turning to LLMs for a quicker, more efficient search process.

The AI Reality

ChatGPT tells a different story. Instead of ten blue links, I get four recommendations. But here's what caught my attention: when I asked the same question on different days, the answers changed. Different banks and credit cards were recommended across different sources. While Google's rankings remain mostly steady and predictable, AI gives each user a unique answer, making the landscape far more dynamic and unpredictable for brands.

The Test

With the help of Scrunch, I designed an experiment to understand what was really happening beneath the surface of AI search. I created 15 different prompts on the theme of "best travel credit cards" to mimic how real people might phrase their searches.

I then simulated these prompts across 4 different LLM models daily until I had completed 500 total searches. This sample size allowed me to see past the randomness and identify genuine patterns in how AI models recommend credit cards and which sources they cite.

The Results

Here's how often each credit card issuers appeared in AI search:

• Chase: 95%
• American Express: 79%
• Capital One: 77%
• Citi: 34%
• Wells Fargo: 16%
• Bank of America: 13%
• Discover: 5%
• U.S. Bank: 2%
• Barclays: 1%

For anyone familiar with travel credit cards, these results probably aren't surprising. The real revelation wasn't which brands appeared most often, it was the sources behind those recommendations. While Google consistently offered me the same 10-15 domains, my AI search results cited over 160 unique websites.

Yes, the usual suspects like NerdWallet and The Points Guy appeared, but so did countless smaller brand websites and niche media outlets such as Thrifty Traveler, Fang Wallet, Finance Buzz, Card Ratings, and Two Dusty Travelers. These smaller players would never crack Google's page one without significant SEO investment, but they did appear in AI search.

What This Means for Publishers

LLMs are leveling the playing field in a way Google never has. It's no longer just about domain authority, backlinks, or how much you've invested in SEO. If you have quality, citable content, you have a chance to appear in an AI answer, regardless of your site's traffic or brand recognition. The gatekeepers are changing. And the gates are opening wider.

Why This Matters for Brands

As more consumers migrate to AI search, brand visibility is no longer just about what you say on your own website. It's about what others are saying about you across the internet. In the Google era, you could optimize your site and control your message. In the AI era, your brand reputation is being shaped by every review, comparison article, and mention scattered across hundreds of sources, many of which you've never heard of.

The brands that understand and monitor their AI visibility today will be the ones that maintain relevance tomorrow. Because in this new search landscape, you don't just need to be found, you need to be recommended. And that recommendation is built from a chorus of voices, not a single website.

Want to discuss AI search and what it means for your brand? Send me a DM or email me at Jpenn@brodeur.com.