.svg)
“How do I rank in ChatGPT?” It is the most frequently asked question in e-commerce at the moment. Closely followed by: “Where do I upload my product feed?” and “Is there a dashboard for this?”
🎧 Listen to the full episode here.
The short answer is: there is no feed upload. There is no dashboard. And ranking in an LLM is not the same as ranking in Google. But the work you do there is remarkably familiar.
Sander Mangel, Solutions Architect and founder of Open Commerce, spoke to Martin Harris, Head of Digital at Tank, a British agency, to cut through the noise surrounding generative engine optimization. Their conclusion: GEO is not a new field. It's the field that most companies should have followed all along.
What GEO actually is (and isn't) GEO stands for generative engine optimization: the practice of making content visible to major language models such as ChatGPT, Claude, and Gemini. The industry treats it as a separate category. Harris disagrees. “If you go through the list that people usually check off as GEO; freshness, original data, structured markup, comparison tables, those are essentially just good SEO principles,” he says. “Here's what people would recommend anyway.”
The difference lies in how LLMs process that content. Google has had 28 years to build an advanced crawling and ranking machine. LLMs are still behind. They are, in Harris' words, “unrefined” in how they retrieve information. They want content that is easy to process, loads quickly and is clearly structured.
A New York Times study with AI startup Oumi showed that Google's own AI Overviews makes mistakes about 10% of the time. At the scale of more than five trillion queries per year, that translates to tens of millions of incorrect answers per day, and if Google still struggles to do so after nearly three decades, the newer LLMs will have an even harder time.
That makes your task clearer, not more complicated: create content that they can read without friction.
Why brand trust is the real ranking factorAhrefs research shows that the vast majority of URLs that ChatGPT cites don't appear in Google's top 10. This suggests that traditional search ranking alone does not determine how visible you are in LLMs. What then?
trusting
“To make recommendations, an LLM must have an element of trust,” Harris explains. “That means a good reputation — in the form of reviews, a strong brand across multiple platforms, or authorship from people who are experts in their field.” Smaller brands are tempted to seek quick results: best-of lists, comparison pages, FAQ blocks. They can work. But they work better when they're backed by a real market position.
“If you're the best in the industry and you can back that up with data, reviews, and testimonials — why not create a list to prove that? That's genuine,” says Harris. “The grey area is where people put themselves at the top based on subjective opinion.” The opportunity for SMEs is real. Harris has seen smaller clients appear in LLM-generated comparisons alongside industry leaders. But differentiation is what gets you there. Generalist brands that try to be everything to everyone are losing out to specialists who own a specific niche and can prove it.
Where to build that reputation
LLMs don't just crawl your website. They draw from media publications, Reddit, LinkedIn, and other platforms where your brand or your people show up. Harris describes a practical approach:
Media and PR: Publications that allow LLM crawlers in their robots.txt file give you a better chance of being cited. Check this before you pitch.
LinkedIn: Original data posted from personal profiles linked to a company page is increasingly being cited by LLMs.
Reddit: Still referenced a lot, although brand building requires genuine contributions, not marketing.
Authorship: Dedicated author pages, bios, and consistent thought leadership across multiple platforms all reinforce the trust signals that LLMs pick up.
“People buy from people,” Harris notes. “A lot of people are dropping out of brands on LinkedIn. It's in a company's best interest to make employees ambassadors and give them a voice.” This is not new advice. But the reason it matters now is that LLMs reward what traditional search engines should have always rewarded: a true brand presence on multiple channels, supported by real expertise.
The measurement problem
There is currently no tool that accurately measures LLM visibility. Everyone who consults an LLM receives a personalized response that is shaped by conversation history and previous interactions. Keeping track of specific prompts is, as Harris puts it, “impossible.” Tools like Peec, LLM Refs and others provide an indication, not a fact.
“This tells you approximately how often you seem to be quoted compared to your competitors. The data isn't accurate,” says Harris. The most reliable approach at the moment is server-side tracking. Microsoft Clarity, which is free, can show which LLM bots visit your pages and how often. If a landing page is visited regularly by bots and frequently quoted elsewhere, that is the closest thing to a reliable signal. The measurement gap is real. But that should not be a reason to wait and see. The fundamentals are clear.
The conclusion
GEO is not a revolution. It's SEO, UX, PR, social and brand reputation that work together. What they should have done all along. “GEO and LLMs and even Google today are much bigger than just SEO,” says Harris. “It's UX, it's social, it's PR, it's digital PR, it's the wider brand reputation you're building around the web. That's where the mind shift must take place.”
The tools will improve. The LLMs are getting smarter. The brands that invested in content over shortcuts are the ones that are still visible when the time comes.
