Shift one: LinkedIn is the fastest-growing citation source
The clearest change in this quarter's data is LinkedIn. Profound's citation study, built from 680 million AI-answer citations collected between June 2025 and May 2026, puts LinkedIn at #1 among cited domains for B2B and software questions across every major engine. Semrush's 325,000-prompt study measured LinkedIn appearing in 14.3% of ChatGPT Search responses.
The detail that matters for anyone acting on this: the engines prefer different LinkedIn surfaces. Perplexity leans on company pages, while ChatGPT and Google's AI Overviews cite individual posts. A brand that only maintains a company page is invisible to half the engines, and a founder who only posts personally is invisible to the other half.
Shift two: citations have decoupled from rankings
Ahrefs tracked 863,000 keywords and found that only 38% of the pages Google's AI Overviews cite rank in the top 10 for that query, down from 76% seven months earlier. Roughly a third of cited pages do not rank in the top 100 at all.
This is the most strategically important number in the report. It means the old playbook, rank first and visibility follows, no longer describes how AI answers are built. The engines are reading conversations and sources well beyond page one. Being present in the specific threads, posts, and pages an engine reads now matters more than outranking anyone on the query itself.
Freshness compounds this: pages cited by AI engines are on average about a year fresher than pages that merely rank, and roughly 90% of AI crawler hits land on content less than three years old.
What has not changed: Reddit is still the widest door
Reddit remains the most-cited domain overall across ChatGPT, Perplexity, Gemini, and Google's AI answers. On Perplexity the concentration is extreme: independent measurements put Reddit at 20 to 24% of all citations, the highest single-domain share on any engine. Nearly all of those citations point at individual threads, not subreddit homepages.
One honest caveat we share with clients: citation volume and brand recommendations are different things. A 36,000-query study by FirstPageSage found community sources behind a small minority of the actual brand names AI engines recommend for buying-intent questions. Reddit gets the engines' attention; what gets your brand named is being genuinely present in the conversations they read. That distinction is the entire reason our placements are real comments in real threads rather than volume for its own sake.
How the engines differ right now
Per-engine citation personalities, from Similarweb and Ahrefs data published in the first half of 2026:
| Engine | Leans on | What that means |
|---|---|---|
| ChatGPT | Wikipedia (~13%), Reddit (~12%), LinkedIn posts | Encyclopedic sources plus community threads; individual LinkedIn posts get cited |
| Perplexity | Reddit (20-24%), tier-1 news, LinkedIn company pages | The most Reddit-heavy engine by far; company pages beat personal posts |
| Google AI Overviews | YouTube (~21%, #1 domain), then web pages regardless of rank | Video answers surged; cited pages often rank nowhere for the query |
| Grok | Reddit (~16%), X posts | Community-heavy with a native X bias |
What we're changing
Three changes to how we run client programs, effective this month.
First, we are weighting LinkedIn placements more heavily in the channel mix for B2B clients, and matching the surface to the engine: company-page presence for Perplexity, individual posts for ChatGPT and AI Overviews.
Second, we are prioritizing freshness. Because cited content skews recent, a steady cadence of new placements beats a large one-time batch, and we schedule accordingly.
Third, Reddit thread targeting continues unchanged. It is still the widest door into AI answers, and the threads that rank on Google for buying-intent questions remain the highest-value placements we can make.