AI Search Strategy · Cross-Functional
PR × AEO — making earned media the engine of AI visibility
In AI answers, most of what gets cited is earned media, not the content you own. I built the strategy and the cross-functional operating model to connect PR and Search so the brand gets named, not just used.
The insight
84%
of what AI engines cite is earned media. Owned and paid together are just 16%.
89%
of B2B buyers consult a generative AI model during vendor research; 42% start inside an LLM.
Many×
higher conversion than organic or paid — but the volume is suppressed when the brand isn't cited.
Source: Muck Rack, "What Is AI Reading?", May 2026.
The challenge
AI answer engines lean on third-party, earned sources — journalism, analysts, reviews, expert commentary — to decide which brands to cite, compare, and recommend. My company had strong owned and AEO content, but PR ran disconnected from Search. The result was a brand that AI leaned on as a data source but rarely named, and a media list that had never been audited against the outlets AI actually cites.
The diagnosis
This was a coordination and targeting gap, not a budget or capability gap. PR and Search were chasing the same outcome from opposite ends with no shared process and no shared targets. The journalists PR pitched and the journalists AI actually cites barely overlapped, because the list had never been built from citation data.
The plan
No new budget, no new headcount.
- 01Close the handoff loop.A recurring PR to Search Visibility process: AEO review before releases go live, and coordinated amplification inside the post-publish citation window.
- 02Retarget the media list.Rebuild pitch targeting around the authors and outlets AI engines actually cite.
- 03Unify expert identity.One consistent person-entity per expert across bylines, on-site author pages, Person schema, and LinkedIn, so AI can attribute expertise correctly. Measured in citation share and brand mention, not impressions.
Status
Presented to marketing leadership and now in motion. PR is engaged, and the shared operating model is taking shape through a recurring working cadence.
What it shows
I find the highest-leverage lever, back it with data, and build the cross-functional operating model to capture it, leading across teams I don't own.