The PR Industry Just Solved the Wrong Measurement Problem
Signal AI's acquisition of Memo confirms what every CCO already felt: traditional PR metrics are broken. But the solution they're building still measures the wrong reader.
Signal AI acquired Memo today.
The pitch: PR teams have been stuck measuring "potential reach" and impressions — vanity metrics that tell you nothing about whether anyone actually read the article. Memo fixes that. Direct publisher readership data, at the article level. Fortune 50 brands like Google, Pepsi, and Walmart already use it. Finally, you know how many humans saw the piece.
It's a real problem. The acquisition makes sense. And it's still optimizing for a reader who is no longer the first one in the room.
Who reads your earned media first now
Here's what Signal AI's announcement doesn't say: the most consequential audience for a Forbes placement in 2026 is not the CMO who skims it over coffee. It's the AI engine that indexes it within hours and starts citing it in answers to your buyers' research queries.
When a VP of Engineering asks Perplexity "what's the most credible [software category] vendor," the answer doesn't come from a human editor who remembered reading your press coverage. It comes from a retrieval system that has indexed your earned media footprint — and decided whether the publications you've appeared in are ones it trusts enough to cite.
That decision happens before any human reads anything.
The numbers are unambiguous. Muck Rack's analysis of over 1 million AI-generated responses found that 85% of AI citations come from earned media — not brand websites, not paid content, not social posts. A Fullintel-UConn academic study presented at IPRRC in February 2026 found that 47% of all AI citations came from journalistic sources, with 95% of cited links being unpaid. Ahrefs confirmed that 65% of ChatGPT-cited pages come from domains with DR80+.
The earned media placements PR has always been trying to prove the value of are the primary input for AI-mediated brand discovery. That is the ROI case. And most PR teams are still building dashboards that don't measure it.
The measurement gap no one is filling
The Signal AI/Memo deal is the latest move in a crowded sprint to fix PR measurement. BrightEdge launched AI Hyper Cube to track brand presence in AI-generated answers. Microsoft added an AI Performance Dashboard in Bing Webmaster Tools this week, letting brands see how their content is being cited across what they're calling the "AI web." Semrush rebranded its entire platform around AI visibility in March 2026. The category is moving fast.
What most of these tools measure: whether your brand appears in AI-generated answers, and how often.
What none of them explain: why some brands are consistently cited and others aren't — and what the actual mechanism is for changing that.
The answer isn't a dashboard setting, a structured data implementation, or a content format shift.
It's the editorial relationships that got your brand into the publications AI engines already treat as authoritative sources. That's the input. The citation is the output. The dashboard just shows you the score after the game is already played.
Why readership data misses the point
Memo's core pitch is sound: if you don't know how many humans read the article, you can't prove PR's value to a CFO. True. But the framing assumes human readership is still the primary signal worth measuring.
Consider what actually happens when a placement goes live in a tier-1 publication today:
- AI crawlers index the piece within hours
- The publication's domain authority signals to retrieval systems that this source is credible
- The specific claims, names, and topics in the article become part of the citation graph those systems draw from
- When buyers research your category, your brand appears — not because a human remembered the article, but because a machine indexed it
Readership and citation authority are not the same thing. A piece in Forbes that 200 people read carries more AI citation weight than a piece in a niche trade publication that 20,000 people read. The Memo data would show the niche piece winning on readership. The AI would be citing the Forbes piece.
This plays out across the data. A 2025 Profound analysis found that only 6.82% of ChatGPT's top citations overlap with Google's top 10 organic results — meaning the publications AI engines draw from are largely separate from the publications that drive the most human web traffic. The Moz 2026 analysis of 40,000 queries confirmed that 88% of Google AI Mode citations are not in the organic top 10.
The AI and human audiences are reading from different lists. Most PR measurement frameworks still assume they're the same list.
Two audiences, two metrics
The PR industry is not wrong that measurement was broken. It has been broken for decades. AVE was fiction. Potential reach was noise. Impressions were a number you showed at board meetings because you didn't have anything better.
Signal AI and Memo are fixing the human readership measurement problem. That matters. But it sits inside a larger structural shift:
| Metric | What it measures | Primary audience |
|---|---|---|
| AVE / potential reach | Nothing useful | Internal budgeting theater |
| Human readership (Memo) | Who actually read the article | CCO → CFO accountability |
| AI citation frequency | Which brands are being recommended by AI systems | Buyer shortlist formation |
| Share of citation | Brand's relative AI visibility vs. competitors | Category authority measurement |
The Memo acquisition closes the gap between rows 1 and 2. Rows 3 and 4 remain largely unmeasured — and they're the ones that determine whether your brand is on the shortlist when a buyer asks ChatGPT who to evaluate.
The Bain and Dynata study of 1,100 consumers found that 80% of search users now rely on AI summaries at least 40% of the time. Forrester research found that 70% of B2B buyers complete most of their research before contacting sales. These two facts together describe a world where AI systems are doing the first cut of vendor qualification — long before any human reads a bylined article.
The readership data tells you who read the piece. The citation data tells you whether the piece got into the research process at all.
Related Reading
- Fintech PR Strategy 2026: Building Earned Authority Without Compliance Risk
- Machine Relations by Industry: AI Visibility Playbooks for 2026
What founders and growth executives should actually track
If you're responsible for brand authority at a company that sells to other businesses, the measurement question that matters is: when your buyers ask AI systems who leads your category, are you in the answer?
Test it now. Open Perplexity or ChatGPT. Ask: "What are the most credible [your category] vendors for [your buyer's use case]?" If your brand isn't named, the problem isn't your readership data. The problem is your citation infrastructure — the earned media footprint that AI engines draw from when they construct that answer.
Building that infrastructure requires earned media in publications AI engines treat as authoritative. Not optimization software. Not measurement dashboards. Relationships with editors at the publications that carry the domain authority to move the needle on AI citation.
That's the mechanism Machine Relations was built to systematize: earned authority in Tier 1 publications → machine-readable trust signals → AI citation → brand discovery before the buyer ever visits your website.
Signal AI and Memo are solving a real problem. The next acquisition in this space will be by someone who understood that the more consequential problem was getting the placement in the first place — in publications that earn the citation.
That's the work AuthorityTech has been doing for eight years.