AI in measurement: Counting, connecting, and getting attention in the algorithm

The world of visibility is changing fast. As large language models (LLMs) like ChatGPT become new information gatekeepers, PR is no longer just competing for audience attention. It’s competing for algorithmic inclusion.

In a digital landscape increasingly shaped by automated content, the quality, credibility, and authority of earned media have never mattered more. What cuts through now is not simply how much content exists, but which content is trusted enough to be cited, surfaced, and amplified by both humans and machines.

At the same time, the way we measure communications is evolving just as rapidly. Our Head of Insights Amy Chappell recently attended the AMEC AI Day, and one message stood out: AI isn’t here to replace human intelligence – it’s here to enhance it. Measurement professionals are no longer just counting the past; we’re connecting data to insight, outcomes, and influence in ways previously impossible at scale.

Together, these shifts point to a fundamental change in how PR and measurement work hand in hand in an AI-driven world: Credible storytelling fuels visibility, and intelligent measurement proves its impact.

Why earned media still dominates

Research presented at the AMEC AI Day suggests that around 90% of AI visibility comes from earned sources, not paid placements. That’s because LLMs favour content that is accessible, credible, and editorially independent. Paid content often falls short on two counts:
It sits behind paywalls or sponsorship disclosures, reducing citability.

It lacks the credibility signals that LLMs prioritise when determining trusted sources.

In an era when 60% of Google searches end without a click, visibility increasingly depends on being cited rather than clicked. AI-generated summaries pull from high-authority, earned sources, meaning quality and credibility of coverage matter more than ever.

The new role of qualitative metrics

If AI models prioritise credible coverage, it’s no longer enough to measure volume or sentiment alone. Understanding the authority and influence of sources, and how well your coverage aligns with your strategic narrative, becomes essential to assessing impact across both human and AI audiences.

In a landscape where automated content is multiplying, human-authored, well-sourced journalism carries greater weight.

That’s why the focus must shift from volume to value: not how many pieces you secured, but how credible, contextual, and influential those pieces are.

Emerging ideas like ‘share of answer’ (which explore how brands appear in AI-generated responses) hint at where measurement might go next. But these are still early indicators, not yet established metrics.

How we measure in the age of AI

Metrics like share of search and Generative Engine Optimisation (GEO) scores are early attempts to quantify visibility in AI environments. But as discussed at the AMEC AI Day, the industry is still testing and calibrating what “good” looks like.

The takeaway? Don’t measure for measurement’s sake.

Storytelling still drives machine understanding

AI is reshaping how visibility works, but not what makes it valuable. The best route to long-term visibility, with both audiences and algorithms, remains the same: authentic earned media, built on credible storytelling, relationships, and expertise.

As the line between human and machine audiences blurs, PR’s superpower endures, creating messages that are not only seen and read, but also trusted.

Where AI supports measurement

AI can assist across every stage of the workflow:

Collection and cleaning: unifying messy inputs from multiple sources.

Categorisation: speeding up tagging and sentiment analysis while ensuring consistency across languages.

Insight generation and prediction: highlighting emerging risks, narratives or audience shifts earlier.

AI’s strengths are clear: speed, scalability, consistency, and cross-market comparability. But its weaknesses are just as important to understand: opaque decision-making, bias in training data, false confidence in generative summaries, and the temptation to switch off human critical thinking.

That’s why we will see a shift from analysts acting less as ‘data producers’ and more as ‘insight curators’, allowing us to spend more time understanding, interpreting, and recommending than ever before. New skills are emerging: prompt engineering, validation, ethical reasoning, and bias checking. These sit alongside the fundamentals: empathy, relevance, and context.

Human accountability remains essential in measurement

Governance is catching up fast. AMEC is developing standards to ensure ethical use of AI in measurement. But the guiding principle is simple: AI can enhance, not replace, human judgment.

The best measurement programmes will use automation for efficiency, freeing up analysts to focus on interpretation, storytelling and strategy. The industry is shifting from manual counting to intelligent contextualisation, and AI is the accelerator helping us get there.

Preparing for the next era of visibility

AI is not a passing trend in communications and measurement; it’s a structural shift in how visibility, influence, and trust are created and understood. For PR teams, that means doubling down on what machines can’t replicate: credible relationships, meaningful narratives, and human judgment. For measurement professionals, it means evolving from trackers of activity to interpreters of influence.

The organisations that will lead in this next era will be those that combine high-quality earned media with intelligent, accountable use of AI, using technology to go faster and further, without losing sight of strategy, ethics, or impact.

Want help with measuring the success of your campaigns? Find out more about Vuelio Insights.

For more about the impact of AI tools on the media and measurement spaces, check out key takeaways shared at the 2025 Press Gazette Future of Media Technology Conference.