LinkedIn profile pages made up 33.9% of the LinkedIn content ChatGPT cited in mid-November 2025. By mid-February 2026, that had fallen to 14.5%. Profound, which tracks AI citations across six models, published the numbers on 9 March 2026, drawing on 1.4 million citations gathered across that three-month window. The profile page is being cited less. Something else took its place.
Posts and articles took the share profiles lost
LinkedIn posts in the main feed rose from 20.9% to 26.0% of ChatGPT's LinkedIn citations over the same three months. Long-form articles rose from 6.0% to 8.9%. Combined, posts and articles went from just over a quarter of LinkedIn citations to over a third - an 8-percentage-point move in three months, the largest shift Profound recorded this year. Over the same window, LinkedIn's overall rank among the domains ChatGPT cites most moved from roughly 11th to roughly 5th, more than doubling its citation frequency.
The mechanism is straightforward once you see it. An AI model answering a question needs a specific line to quote or paraphrase: a stated view, a number, a named example. A profile page is a summary written in the third person. A post is someone making a claim in their own words. One of those is easier to lift a sentence from, and the citation data is now moving the way you'd expect.
This builds on a pattern already visible in the data
Semrush's and OtterlyAI's earlier analysis of LinkedIn citations already found named individual authors taking the large majority of citations, well ahead of company pages and reshares. Profound's newer numbers don't contradict that. They show the same advantage getting stronger each quarter it's been measured. The gap between a maintained personal profile and an active posting habit is opening up, not levelling off.
What this changes about where the effort goes
A complete, well-written profile is still worth having. It's a reasonable baseline, and it's increasingly not the thing doing the citing. The content type gaining ground is the individual post or article, published under a real name, saying something specific enough for a model to quote directly. Three months is a trend, not a permanent rule. Profound scopes its own finding to that window and makes no claim about what happens next. But the direction over that window is single, not mixed. The profile page still matters, but it's not where the growth is happening. The posts and articles underneath it are.
A few common follow-up questions
Does this mean a company LinkedIn page is a waste of time?
Not a waste, since it still gives an AI model context about who a person works for. But it isn't where the citation growth is. Semrush's and OtterlyAI's studies both found named individuals taking well over 90% of LinkedIn citations, against single digits for company pages.
How often is this worth checking again?
Quarterly is a sensible interval, given how fast Profound's own numbers moved. Checking more often than that mostly picks up noise. Checking less often risks missing a genuine shift like this one until it's already old news.
Does this pattern apply outside LinkedIn?
Profound's data covers LinkedIn specifically. The underlying logic - a specific first-person statement gives a model something to quote, a generic bio doesn't - probably holds wherever a model is choosing what to cite, but no comparable dataset exists yet for other platforms to check that against directly.