The GEO metrics nobody is tracking yet

Traditional SEO reporting still revolves around the same familiar numbers: rankings, impressions, clicks and traffic. Although useful, these numbers don’t give the complete picture anymore. 

AI search is changing where visibility happens and a growing share of that visibility never turns into a measurable website visit.

A potential client might discover your brand through ChatGPT, compare suppliers in Perplexity, skim a Google AI Overview and only visit your site days later. Some users may never click through at all.

Generative engine optimisation (GEO) is about helping AI-powered search tools understand, trust and surface your brand in conversational answers. But if search behaviour is changing, the way we measure performance has to change too.

Traditional SEO metrics aren’t broken, they’re incomplete

Let’s be clear, rankings, clicks and traffic still matter and SEO is not being binned because AI Overviews exist. But those metrics were built for a search environment where the user typed a query, scanned a results page and clicked a blue link.

That journey is no longer guaranteed.

Ahrefs found that Google AI Overviews reduced organic click-through rates (CTR) for position-one results by 58 per cent in a recently updated study. Seer Interactive data, reported by PPC Land, found organic CTR for informational queries featuring AI Overviews fell 61 per cent and paid CTR fell 68 per cent.

So a page can rank well and still lose traffic. A brand can be mentioned in an AI answer and never receive the click, while a buyer can be influenced by your content without becoming an obvious organic session.

That doesn’t mean performance has vanished. It means your dashboard may not be seeing it. Search is becoming less about links and more about answers.

Here are the six GEO metrics you should start paying attention to now.

1. AI brand mentions

The first GEO metric to track is simple: how often does your brand appear in AI-generated answers?

This matters because AI search is not just returning pages, but making choices. It summarises, recommends, compares and filters. If your brand appears in those answers, you’re part of the conversation. If it doesn’t, you may be invisible even if your SEO rankings still look healthy.

The key is to track mentions across prompts that are genuinely relevant to your business, products or services. That might include branded searches, category searches, comparison queries or common customer questions.

For example, a project management software company might monitor prompts such as:

• Best project management tools for remote teams
• Alternatives to Asana for small businesses
• Which software is best for managing marketing workflows?
• How do I improve team collaboration across departments?

And it’s important to check across multiple platforms, as ChatGPT, Gemini, Claude, Copilot and Perplexity all behave differently. These models pull from varying sources, interpret authority differently and prioritise different content structures.

The point is not to obsess over one perfect prompt. It’s to understand whether your brand is consistently recognised in the kinds of conversations your audience is already having.

2. Share of AI voice

In traditional search, marketers track keyword visibility, rankings and share of search. In AI search, the equivalent is the share of AI voice. This measures how often your brand appears in AI-generated answers compared with competitors.

This matters because AI search compresses choice. A traditional SERP might show ten organic listings, ads, videos, forums and “people also ask” results. However, an AI-generated answer may recommend three brands, sometimes fewer.

In AI search, second place can disappear quickly. For example, if ten high-intent prompts generate 100 total brand mentions and your brand appears 12 times, your share of AI voice is 12 per cent. It’s not a perfect metric, but it gives you a useful directional view of your visibility within AI-generated conversations.

Consider tracking:

• Competitor mentions
• Recurring recommendation patterns
• Category-level visibility
• Branded vs non-branded prompts

Over time, this helps show whether your SEO and GEO strategies, content and digital PR activity are increasing your authority in AI search environments.

3. Citation frequency

Citation frequency tracks how often AI tools cite or link to your content as a source. This is especially important in Google AI Overviews, Perplexity and other systems that show source links.

It helps answer questions such as:

• Which pages are AI tools citing?
• Which topics generate the most citations?
• Are competitors being cited more often?
• Are citations coming from your site or third-party mentions?
• Are AI tools using your strongest, most accurate content?

Google is also evolving how links appear inside AI search. Search Engine Journal recently reported that Google is introducing more inline links and contextual citation features in AI Mode and AI Overviews.

That matters because it suggests source visibility is important even as search becomes more answer-led. Being mentioned is good, but being cited as the source of the answer is better.

4. Prompt visibility

Keywords still matter, but they’re no longer the only entry point. AI search is conversational – users ask full questions, refine their searches and add context naturally.

That means GEO reporting needs to track prompts, not just keywords. Prompt visibility measures whether your brand appears across broader natural-language searches around your topic.

Where a traditional keyword cluster might include:

• GEO agency
• AI search optimisation
• Generative engine optimisation

A conversational search expands far beyond that:

• How can my brand appear in ChatGPT recommendations?
• What should I track for AI search visibility?
• Is GEO different from SEO?
• What content formats are easiest for AI search tools to cite?
• Which agency can help with AI search strategy?

AI search strategy refers to this as “query fan out” –  the web of related questions users ask AI tools during research and discovery.

That means brands need visibility across themes and conversations, not just isolated keywords.

5. AI referral traffic

AI referral traffic is the metric most teams already want to see. It tracks visits from platforms such as ChatGPT, Perplexity, Gemini or Copilot.

It’s useful, but it should not be treated as the whole story. AI referral traffic is often small, messy and under-attributed. Some AI interactions don’t pass clean referral data. Some influence happens without a click and some buyers may ask AI for advice, then later return via branded search or direct traffic.

AI referral traffic is one signal worth measuring, but it’s not the only one that matters.

Track it, but pair it with other indicators such as branded search demand, assisted conversions, sales feedback and prompt visibility. Otherwise, you risk underestimating AI’s role simply because the user did not click in a neat, trackable way.

6. AI sentiment

Visibility is only helpful if the answer is useful, accurate and favourable.

AI sentiment tracks how your brand is described in AI-generated responses. Are you positioned as credible? Specialist? Expensive? Generic? Out of date? Are competitors described more clearly?

This matters because AI answers increasingly shape perception before someone reaches your website. Business Insider reported research suggesting Google AI Overviews were more likely to talk negatively about brands than ChatGPT in the data reviewed.

Even when the data vary by sector, the principle stands that AI search is becoming a reputation layer. You need to know not just whether you appear, but how you appear.

To determine this, track:

• Tone of mentions
• Recurring descriptions
• Inaccuracies
• Trust-related language
• Competitor positioning

Because visibility without trust is not much use.

What should marketers report now?

A useful GEO report doesn’t need to replace your SEO, but should sit alongside it. A practical monthly GEO dashboard could include:

• AI brand mentions
• Share of AI voice
• Citation frequency
• Prompt visibility
• AI sentiment
• AI referral traffic
• Branded search trends
• Competitor visibility
• Top cited pages
• Content gaps by prompt cluster

This gives you a more honest view of visibility. It moves the question beyond ‘Did we get traffic?’ to ‘Are we being recognised, referenced and recommended where people now search?’. 

Where Axonn can help

We help our clients adapt to AI-powered search by combining:

• SEO
• GEO
• Content strategy
• Authority building
• AI-friendly content optimisation

That means creating content AI tools can actually understand, trust and cite – while still producing useful, human-first experiences.

It also means connecting SEO, content strategy and analytics. GEO isn’t one team’s responsibility anymore. It sits across your entire digital presence.

The future of search is not just about rankings, but about becoming part of the answer. Contact us to learn how we can help you become more AI visible.