PR Measurement in the Age of AI Search

As a communications leader, you measure what matters: website traffic, share of voice, media impressions. Earned value equivalencies. These metrics have shaped how PR teams justify their work for decades.

But now in the age of LLM search, there’s a new channel to track for brand visibility.

When someone wants to learn about your brand before a pitch meeting, purchase, or partnership, they're increasingly turning to an LLM rather than searching Google. What that answer says about you, and which of your coverage is driving it, has been almost entirely invisible to PR and comms teams. Many PR & comms pros are unsure how to measure it, whether the best stories are the ones the AI cites, or whether an old crisis report carries more weight.

This is exactly what we explored in a webinar with Eddie Kim, President of Memo, and Charlie Baldwin, SVP of Insights & Analytics at We. Communications. What they revealed is that PR's measurement problem in the AI era isn't a data problem. It's a problem of visibility, scalability, and ownership.

GEO Measures the Wrong Thing

Ask a PR director what's showing up in AI answers about their brand. Most will pause. Some will mention they've typed a few questions into ChatGPT. Few will have a systematic answer or reliable tool, and that's become a competitive disadvantage.

The industry's response to AI search or LLM visibility has been Generative Engine Optimization, or GEO. The logic seems sound: if LLMs are the new front page, optimize your content for them by updating your website, restructuring your blog, adjusting your metadata, and measuring your owned content performance.

Here's the gap: industry data show that 80-90% of what LLMs cite comes from earned media rather than owned content. GEO, as currently practiced, measures and optimizes the 10-20% minority and says nothing about the coverage actually shaping what AI says about you.

During the webinar, Charlie Baldwin pointed out that:

"[With] journalism being the single most important, most influential source of influence in AI answers… communications professionals are indicating that they need to have a seat at the table, they need to be leaders in this space, because there's a lot of truth to that."

If 80-90% of AI visibility comes from your earned media, then your earned media is what you need to measure. GEO is a solution built for marketers. But measuring the impact of earned media in AI answers? That's a communications problem to tackle.

Your PR Metrics Don't Tell You What Matters Anymore

Reach isn't the problem, relevance is. An article can have a massive human readership and still be invisible in AI answers. Or it can have modest reach and be the primary source the AI cites when answering questions about your brand.

Baldwin also explained how this plays out in practice:

"It might be that the LLMs prefer some of the smaller trade publications or some of the smaller publications that have more direct and more complete answers to the questions that were posed. That might lead you to a different marketing or a different comm strategy."

This is the measurement gap. Your biggest, most prestigious placements might not be what the AI cites. The publications you've historically deprioritized might be the ones shaping what AI says about you, and your traditional PR metrics (reach, impressions, and earned value) won't tell you any of this.

If that includes a crisis report from three years ago or an unchallenged competitor framing, the AI will cite it with the same authority as your latest feature story. You can't manage that risk if you can't see it.

This is where your measurement problem becomes acute. You need to know which stories actually move the needle in AI-generated answers, which publications the models trust, which narratives are shaping what AI says about you, and you need to track it over time.

How to Close the Measurement Gap

The data exists. What's missing is visibility into it.

In the webinar, Baldwin advised the audience to run a measurement audit:

• Define 10 key questions about your brands
• Simulate searches across multiple LLMs
• Collect two weeks of data, studying both the inputs and outputs

From there, you can see the patterns. Which publications are the LLMs citing most? Where is your narrative winning, and where are you invisible?

One example: we analyzed a credit card brand that appeared in only 33% of AI conversations. The comms team was targeting broad national outlets, the outlets that traditionally drive reader reach. But LLMs were actually citing niche fintech publications they weren't actively tracking. Armed with that measurement, the strategy shifted. Instead of chasing headlines everywhere, they targeted the specific reporters whose publications the AI actually cited.

That's PR measurement built on data that didn't exist a year ago. This isn't about guessing or following trends. It's about measuring what you can actually control and building out a strategy from there. It's about knowing your earned media impact in a new context.

Now You Can Measure It

The most consequential shift in search isn't that LLMs replaced Google. It's that systems are now forming opinions about your brand at scale, drawing on your coverage history and synthesizing it into answers that shape perception.

Every journalist, every publication, every placement decision is now an input into what AI says about you. And for the first time, you can measure which ones actually move that needle.

Signal AI's AI Citations tool reveals exactly which earned media your brand appears in across LLM responses. It shows you your visibility score, which publications the models trust most, which narratives are driving your presence, and where you're getting left out of the conversation entirely. You can benchmark it over time and build a real measurement framework for PR in the AI era.