Someone in my network posted a Google Search Console screenshot last week to prove their “AI search strategy” was driving results. Clicks and impressions from Google. As evidence that their client was winning in ChatGPT.
This is not a one-off. It’s becoming the standard proof point in GEO and AEO pitches. And it’s completely nonsensical.
You cannot rank first on ChatGPT
This needs to be said plainly because an entire consulting niche is being built on the premise that you can: there is no #1 ranking in an LLM. That’s not how the technology works.
LLMs are probabilistic. They generate unique responses. Ask ChatGPT the same question from two devices and you’ll get different answers, different brands mentioned, different sources cited. The whole architecture is designed to not produce a fixed, ordered list.
When someone tells you they can get your brand to “rank #1 on ChatGPT,” they are either misunderstanding the technology or counting on you to misunderstand it. There is no SERP. There is no position tracking. The mental model borrowed from Google simply does not transfer.
The metrics problem is worse than it looks
The Google Search Console issue isn’t just a wrong screenshot. It reveals a deeper confusion about what to measure and why.
Click-through rate from ChatGPT is abysmal by design. ChatGPT is not a list of links. It’s a conversation. Users get their answer in the response. They don’t click through the way they do from a search results page. Measuring CTR as your success metric means you’re optimizing for a behavior the platform actively discourages.
What actually signals LLM visibility looks nothing like traditional search metrics. Self-reported attribution (“How did you hear about us?”). Branded search growth over time. Direct traffic increases. Deep page entries where someone lands on a specific product or comparison page without a referral trail that makes sense through Google alone.
These are harder to track. They don’t fit neatly into a dashboard screenshot for LinkedIn. Which is probably why nobody’s posting them.
What a real strategy looks like
The useful framing isn’t “How do we rank in AI?” It’s “How are LLMs arriving at their recommendations, and where is our brand in that process?”
Run a hundred prompts relevant to your business across multiple devices. Do it more than once. Track how often your brand appears and, more importantly, why it appears. The “why” breaks down into a few patterns:
Some prompts pull directly from your website. ChatGPT cites your content as the reason it mentioned you. This is your owned content doing its job.
Some prompts pull from third-party listicles, comparisons, and directories. This is traditional off-page SEO and link building showing up in a new context.
Some prompts pull from Reddit, YouTube, and social. This is the public conversation about your brand feeding the model’s understanding.
The strategy isn’t a new discipline. It’s understanding how the same fundamentals (content, backlinks, brand presence) get processed by a different system. The companies doing this well aren’t running a separate “AI strategy.” They’re building consensus across touchpoints: their own site, third-party coverage, directories, forums. Then they’re tracking whether that consensus shows up when LLMs generate answers.
Be selective about who you listen to
The amount of misleading content about SEO and LLMs on LinkedIn has hit a peak. People who have never audited how a single LLM response gets assembled are selling “GEO packages.” The proof points are recycled Google metrics dressed up in new language.
If you’re investing in this space, the first filter is simple: ask your potential partner to explain how an LLM decides which brands to mention in a response. If they can’t answer that clearly, or if their answer sounds like “we optimize your content for AI,” keep looking.
The opportunity is real. The measurement framework is just different from what most of the market is selling.