77% of Americans have used ChatGPT as a search tool.
Let that settle for a moment. Not as a writing assistant. Not as a coding helper. As a search tool — the thing they used to do on Google, and before that on Yahoo, and before that by asking someone who knew things.
Three quarters of the American internet-using population has already rerouted their curiosity through a large language model at least once, and the advertising industry is still debating whether AI search is “a thing to watch.”
It is not a thing to watch. It is already the floor. The ceiling is still being built.
The question that matters right now is not whether your audience is discovering information through AI. They are.
The question is whether your brand is visible when that happens — and more importantly, whether you have any idea how to measure it.
How a mattress company accidentally solved the problem first
In early 2024, Casper’s content team noticed something odd.
Their organic traffic from Google had plateaued. Not dropped — plateaued.
But their brand searches were climbing.
New customers were arriving on the site already knowing product details that weren’t prominent on the homepage. When they surveyed those customers, a meaningful cluster said something like: “I asked ChatGPT which mattress was best for back pain and it mentioned Casper.”
The team couldn’t track it.
There was no UTM parameter on a ChatGPT recommendation. There was no referral source in GA4.
It was invisible in the dashboard but visible in the business.
Revenue was there; attribution was not.
So they did something methodical. They started running weekly manual prompts across ChatGPT, Perplexity, and Google’s AI Overview for 40 of their target queries — things like “best mattress for side sleepers under $1500″ and “most durable foam mattress.”
They logged whether Casper appeared, in what position, with what framing, and whether the language used mirrored their own content or a competitor’s.
Within three months they had a proprietary dataset that their competitors didn’t have.
They used it to double down on specific content angles — particularly third-party review signals and long-form comparison guides — that appeared to influence AI responses.
By Q3 2024, Casper’s mention rate in tracked AI queries had improved by roughly 40%.
No one handed them a metric. They invented one.
What AEO actually measures — and why most marketers are flying blind
Answer Engine Optimisation — AEO — or GEO optimization is the practice of making your content, brand, and authority legible to AI systems that generate responses rather than return links.
Think of it as SEO’s more complicated, less documented sibling. Where SEO asks “can Google crawl and rank this page” AEO asks “will an AI cite this source, mention this brand, or adopt this framing when a user asks a relevant question.”
The problem is that the metrics ecosystem for AEO is still being built in real time.
Most marketing teams are either ignoring it or measuring it with tools designed for a different era.
Here is what the field currently looks like, and where the leverage is.
Brand mention rate in AI outputs
This is the most fundamental AEO metric.
It answers the question: when someone asks an AI assistant a question in your category, how often does your brand appear in the response. It is not the same as a ranking.
An AI response is not a list of ten blue links. It is a synthesised answer, and your brand either contributes to that synthesis or it doesn’t.
Measuring this requires manual prompt testing or tools like Profound, Otterly, or BrandRank.AI, which systematically query AI engines and track brand visibility over time.
As Rand Fishkin put it in a 2025 SparkToro newsletter: “The brands that will win the next decade of discovery are the ones that become part of the AI’s assumed knowledge — not just the ones that rank on page one.”
The mechanism here is simpler than it sounds.
LLMs are trained on text. They cite, paraphrase, and defer to sources that appear frequently, authoritatively, and consistently across the web.
If your brand is discussed in depth on high-authority sites — review platforms, industry publications, Reddit threads, independent blogs — the model has more signal to draw from. Brand mention rate is therefore a lagging indicator of your content and PR strategy.
Most marketers get this wrong by treating AEO as a technical SEO problem
Comments