Why Your Brand Might Be Missing from ChatGPT (Even If You Rank on Google): What AI Models Look For

I wanted to see how well AI tools like ChatGPT understand product queries compared to Google, and whether the recommendations align with what ranks organically.

I ran a prompt in ChatGPT asking for recommendations for the best heavyweight oversized t-shirts. Here’s what I found!

ChatGPT vs Google: Two Very Different Lists

When I asked ChatGPT for the best heavyweight t-shirt brands, it returned a structured list:

Premium brands: Represent, Essentials, Aime Leon Dore

Best value options: Uniqlo, H&M, Zara

Honourable mentions: Carhartt, AS Colour, and more

What stood out was how ChatGPT explained its criteria, specifically calling out that a GSM (grams per square metre) rating above 220 was considered “heavyweight.” 

That became a key filtering parameter for its recommendations.

In contrast, when I searched Google for "Surge Heavyweight Oversized T-Shirt", the top organic listings included:

• M&S

• Hera Clothing

• E2Q

• ASOS

Interestingly, none of these brands were mentioned by ChatGPT.

Why Was Hera Clothing Missing?

Hera Clothing ranks highly on Google for the search term in question (usually position 1 or 2).

But ChatGPT didn’t even acknowledge it when listing top heavyweight tees. So I asked it directly:

“Why didn’t you include Hera Clothing in your list?”

ChatGPT flagged that Hera’s products might not meet the “heavyweight” definition due to unclear or missing GSM specs. It also suggested that their products may appear premium-priced without using premium materials, referencing quality concerns or lack of data around product composition.

If ChatGPT doesn’t have clear data points or quality signals about your product, it simply won’t recommend you.

ChatGPT & Rigid Definitions

After running similar tests like this, and asking ChatGPT to explain why it’s included certain products and brands in the list, it’s clear ChatGPT often has a definitive definition based on what you’re asking. 

In some cases, it’s obvious, but sometimes it can be more nuanced. 

In this case, the clear definition for a ‘heavyweight’ t-shirt was over 220gsm. 

I’d always ask ChatGPT how it defines something, and it might help you work out why your brand isn’t getting listed. 

This reinforces the need to define your products in a way that AI can understand, not just for human users. 

Think like a machine: if a spec like GSM is essential to the definition, make it prominent and consistent.

What This Means for Brands and SEO

This is a sign of how search is evolving in the age of AI. Here are three key lessons:

1. Product Data Is Critical

If your product specs (like GSM, material, fit, sustainability, etc.) aren’t clearly marked up on your site, AI tools may not “see” them — and you risk being excluded from recommendations entirely.

2. Brand Perception Matters

ChatGPT factors in brand reputation signals and inferred quality. If your reviews are mixed, or if your product doesn’t stand out with unique or premium attributes, you might be filtered out even if you rank well on Google.

3. We’re Moving Beyond Traditional SEO

Optimising for AI search tools means thinking more like a product marketer. It’s not just about keyword usage - it’s about aligning your offering with the way these tools interpret user intent and quality standards.

Back to Hera & Their Heavyweight T-shirts

After a little look on the Hera heavyweight t-shirt product page (which ranks in position 2 on Google), it became clear from the product description the weight of the t-shirt is actually 328gsm. 

This is much higher than the minimum ‘requirement’ for a heavyweight t-shirt defined by ChatGPT. 

So, I gave ChatGPT this information and asked why it didn’t know about this. 

In short, ChatGPT said it didn’t even come across this page when searching for it, and elaborated that; 

• Hera have traditionally sold t-shirts well under 220gsm 

• Reviews were mixed about the quality of the t-shirts 

• Hera hasn’t always defined the GSM in their product listings

• ChatGPT also elaborated and said the brands it mentioned in the first response often listed GSM and that ChatGPT reviewed this during it’s training (which is fascinating). 

So, even though Hera now lists their GSM within their product description, ChatGPT is yet to pick this up. 

This is a vital thing to consider when we optimise for ChatGPT and other LLMs, as it will still regularly revert back to training data (for now anyway).

Summary & Key Takeaways

• Ensure key product specs (like GSM in this instance) are clearly stated and easy for AI to interpret.

• Recognise that brand perception, quality signals and reviews influence AI recommendations.

• Understand that LLMs often revert to older or incomplete training data, so real-time updates may be missed.

• Ask ChatGPT why they have selected certain brands, or why they haven’t selected other brands - this will give you some good insight. 

As LLMs evolve and retrieval-augmented generation (RAG) becomes more common, the brands that thrive will be those that combine strong on-page SEO with clear, structured, and trustworthy product data.