Measuring GEO isn’t impossible, stop pretending it is

Since launching our GEO service last week, one thing has become clear: people are still confused about measurement. Specifically, there’s a strange narrative circulating among some PRs that measuring visibility in generative AI is either impossible or, at best, unreliable. Let’s be honest: that’s nonsense.

Measurement is possible. It’s already happening. And it doesn’t require building some mystery box of proprietary dashboards to make it work.

GEO isn’t magic; it’s just search, evolving

Generative Engine Optimisation is a new name, yes. But at its core, it’s about visibility in discovery. Just like SEO, it relies on clear signals: structured content, trusted sources, and brand authority. And like SEO, it can be tracked.

There are now industry-standard tools: from long-established players like SEMrush Enterprise’s AI Optimisation suite to newer entrants focused specifically on AI insights, such as Evertune AI. This graphic from a16z alone shows how many services are available and how misguided the “measurement isn’t possible” narrative really.

These tools can show how often your brand is cited across GenAI platforms like ChatGPT, Perplexity, and Google’s AI Overview. With this toolkit, you can track where, how, and in what tone your brand appears. You can see which content is influencing results, the key messages, keywords, and questions involved. You can benchmark against competitors.

Beware the self-made black box

More concerning is the rise of proprietary “GEO measurement platforms”: in-house tools that are rarely peer-reviewed and often vague on methodology. We’ve seen this story before: give your client a dashboard, brand it well, and hope they don’t ask how it actually works. In a discipline already grappling with trust and transparency, this is the last thing PR needs.

If your measurement system can’t be replicated, can’t be interrogated, and doesn’t align with established search metrics, then it isn’t measurement. It’s theatre.

The tools exist. The data is there. The challenge is understanding it.

Our Fusion team isn’t building black boxes at Fire on the Hill. We’re building strategies based on real tools and real data. We track AI visibility across multiple platforms, measure which sources are being cited (Wikipedia, analyst reports, the brand’s blog), and continuously adapt based on what’s working. This is marketing science, not magic.

Of course, GenAI behaviour is inconsistent. Of course, results vary by prompt. But that’s no excuse to shrug and say, “you can’t measure this.” If anything, it’s a reason to double down on rigour, combine multiple tools for greater assurance, and stop hiding behind uncertainty.

So, what should PRs do?

Stop saying GEO measurement isn’t possible. It is. Stop pretending proprietary dashboards equal credibility. They don’t. And stop seeing this as someone else’s problem. If PR wants to lead in this space (and it absolutely should, since coverage and content remain king), then we need to approach GenAI the way we’ve approached media for decades: with clarity, credibility, and a commitment to being the trusted voice in the room.

Article written by David Clare, director & head of Fusion at Fire on the Hill.