Two unrelated datasets put the same platform in front of the two audiences that decide how you get described.
One profile, two sets of readers
Your LinkedIn profile gets read by people deciding whether to work with you. It also gets read by systems deciding how to describe you, and by reporters deciding whether to call.
Those readers want different things from the same few thousand characters, which is the reason most profiles serve one of them well and the others badly.
Where LinkedIn sits in the AI citation data
Profound’s “AI Platform Citation Patterns,” published June 5, 2025 and updated in August 2025, analyzed 680 million citations from August 2024 through June 2025. Its query sampling method is not disclosed, so the shares below describe Profound’s sample rather than the whole web.
Within Google AI Overviews’ top 10 sources, LinkedIn accounted for 13.0 percent. Reddit led at 21.0 percent, YouTube followed at 18.8 percent, Quora at 14.3 percent and Gartner at 7.1 percent.
Within Perplexity’s top 10, LinkedIn took 5.3 percent, behind Reddit at 46.7 percent, YouTube at 13.9 percent, Gartner at 7.0 percent and Yelp at 5.8 percent.
ChatGPT’s top 10 in the same dataset contains no LinkedIn at all: Wikipedia at 47.9 percent, Reddit at 11.3 percent, Forbes at 6.8 percent, G2 at 6.7 percent and TechRadar at 5.5 percent. Which assistant someone uses changes which of your pages answers for you.
A second dataset, a smaller share
LLM Pulse‘s top cited domains list, updated October 26, 2026 on a seven-day rolling window, puts linkedin.com at 2.86 percent, behind youtube.com at 9.73 percent, google.com at 6.25 percent, reddit.com at 5.97 percent, facebook.com at 4.92 percent and instagram.com at 3.33 percent.
The two figures measure different things. Profound’s 13.0 percent is a share within one platform’s top 10 sources, while LLM Pulse’s 2.86 percent is a share across cited domains generally.
LLM Pulse also carries an unresolved caveat on the page itself, since its chart heading says last 90 days while the methodology says seven days, and the sample size is not published. Read it as a second opinion rather than a correction.
Where LinkedIn sits with reporters
Cision’s 2026 State of the Media Report, fielded in January and February 2026 across 19 markets with 1,899 journalists, found 62 percent use LinkedIn professionally and 33 percent rank it as the single most valuable platform for their work.
A third of a professional press corps naming one platform as their most valuable is real concentration, and it reads differently next to the same survey’s finding that 49 percent name shrinking budgets, staff cuts and heavier workloads as their biggest obstacle.
Reporters with less time consolidate onto fewer tools. The tool they consolidate onto is where your profile has to be correct.
Trust is the other half of that number
Muck Rack’s 2025 State of Journalism, fielded in April 2025 with 1,890 journalists and 1,515 responses after cleaning, found 60 percent trust LinkedIn to treat journalistic content fairly. In the same survey, 72 percent found X untrustworthy.
Usage and trust pointing the same direction is uncommon in platform data. A reporter who finds you on LinkedIn tends to treat what they find there as reasonable evidence about you.
The overlap is the useful part
Your profile is one of the few assets that serves both readers from a single file. A model drawing on AI Overviews’ fourth-ranked source and a reporter using their most valuable platform arrive at the same page.
Most other work splits by audience. A press release reaches journalists, a schema block reaches parsers, a conference talk reaches a room.
What that changes about the copy
I run Subscribe PR in Phoenix, and I write the profile as a reference document first and a pitch second.
The headline carries the role, the company and the subject you are a source on, in plain words a model can lift without interpretation. The about section states facts in complete sentences, because fragments parse badly and read worse.
Each claim on the profile also appears on an independent page, so the profile corroborates the rest of the record instead of asserting on its own.
Before you rewrite it
Check what search engines and AI assistants return for your name today, because the profile has to agree with the rest of the record to be worth the effort.
Caleb Hellinger writes about which sources an assistant pulls from when it answers for a name in the write-ups at calebhellinger.com/insights.
Caleb Hellinger runs Subscribe PR, an answer engine optimization and digital PR agency in Phoenix.






