We ran an audit of 63 B2B companies to see where AI answer engines pull their social sources from. One number stood out. When those engines quote a named person, not a company, not a forum thread, a person, that person published on LinkedIn 74% of the time.
It’s our own audit, not a published research study, so treat the numbers as a snapshot of what we found.
B2B buyers are already asking AI answer engines to summarize a category before they talk to a salesperson. Marketing teams have started tracking how often their brand shows up in those answers, the same instinct that drove a decade of SEO work. Most of that tracking focuses on which sources get cited most often overall. Our audit asked a narrower question: when the answer includes a named person’s opinion, where did that person publish it?
AI answer engines don’t just cite sources. They cite people. A model answering “what do B2B marketers think about X” wants a name attached to the opinion. Across the 63 companies in our audit, we counted every named voice the engines quoted outside of Reddit’s community threads. There were 385 of them. LinkedIn accounted for 285 of those citations. Medium was next, at 12.5%. Substack came in at 4.9%. Quora and Hacker News tied at 3.1% each. SlideShare took 1.6%, and X took 0.5%.
LinkedIn wasn’t close on this measure. It was cited for 52 of the 63 companies we audited, 83% of them, more companies than any other network.
What this means for your social program
If your company has a point of view worth quoting, LinkedIn is where AI answer engines expect to find the person holding it. A company blog post can inform a model. A LinkedIn post from your VP of product, your head of sales, or your founder gives the model a name to cite.
This changes how you should think about your social program. A company page builds the audience and sets the record straight on what your brand stands for. But company pages made up only 6.7% of the social citations in our audit. Individual posts by named people made up 32.8%. We dug into that split separately in who AI actually cites, your employees or your company page. Your experts, publishing under their own names, are doing work your company page can’t do on its own.
Think about why a model would prefer a named source in the first place. An answer engine assembling a response wants to attribute a claim to someone accountable for it. A company page speaks for the brand in aggregate. A named post carries a person’s credibility, their job title, their history of posts on the topic. That’s easier for a model to cite as “according to [name], [title] at [company]” than to attribute a general company statement the same way.
The Reddit question
Reddit is the single largest source of social citations in our audit, at 55.8% of all social citations. Almost all of that is community discussion, threads and comments from many contributors rather than a named individual making a claim. That’s a different kind of source than a person publishing an opinion piece. Community discussion answers “what are people saying.” A named LinkedIn post answers “what does this specific person think, and can I attribute it to them.”
Both matter to an AI answer engine building an answer. But if your goal is getting your own experts quoted by name, Reddit’s community threads aren’t where that happens for most B2B companies. LinkedIn is.
The numbers behind this
Some context on scale. Across the 63 B2B companies in our audit, we tracked 27,181 total citations from AI answer engines. Social sources accounted for 2,720 of them, 10.0% of everything cited. Of those social citations, community discussion (mostly Reddit) made up 58.9%, individual posts made up 32.8%, and company pages made up 6.7%.
Social is a minority of citations overall. Most of what an AI answer engine cites still comes from owned content, review sites, and press coverage. But social is the layer where a named person can show up in the answer at all. A blog post gets attributed to the company. A LinkedIn post can get attributed to the person who wrote it, by name and title.
Put those together and the pattern is specific: when an AI answer engine wants a community’s general sentiment, it pulls from Reddit threads. When it wants a named expert’s opinion, it pulls from LinkedIn 74% of the time. Those are two different jobs, and LinkedIn is winning the one that puts your people’s names in the answer.
What to do with this
The practical takeaway isn’t “post more on LinkedIn.” It’s “get more of your named experts publishing on LinkedIn under their own names, on the topics you want AI answer engines to associate with your company.”
A few starting points:
Identify your quotable experts. Who at your company has a real opinion on the topics your buyers ask AI answer engines about? Product leaders, sales leaders, and subject-matter experts are the most citable, not just the CEO.
Publish under their names, not just the company page. Company pages have their place. They’re not where citations concentrate.
Cover the specific questions your buyers ask. A named post that directly answers a real buyer question is more citable than a general thought-leadership piece.
Check where you stand today. Oktopost’s free AI visibility audit shows which AI answer engines already cite your company. It also shows which competitor voices are showing up instead of yours.
An advocacy program is the mechanism
Getting one executive to post occasionally won’t move a citation count. Getting a dozen named experts publishing consistently will.
A structured employee advocacy program is how that happens at scale. It gives your product leaders, sales leaders, and subject-matter experts a steady supply of content to post under their own names, without asking each of them to become a full-time writer. It also gives marketing visibility into what’s actually going out, which matters for compliance and consistency once more than one or two people are posting regularly.
None of this replaces a strong company page or a good content strategy. It sits alongside them. The company page still carries your brand voice and reaches people who follow the company directly. What the advocacy layer adds is the named, attributable voice that AI answer engines are already pulling into their answers three times out of four when a person, not a page, gets quoted.
Run the free AI visibility audit to see where your company stands. It shows which named voices AI answer engines are already surfacing in your category, and where the gaps are against competitors doing this well.
The post When AI answer engines quote a person, that person is on LinkedIn 74% of the time appeared first on Oktopost.
