How to Get AI to Recommend You by Name (In One Afternoon)
You want to be the name that comes up when someone asks about your thing. Not your company's name, yours.
That used to take a decade of speaking gigs and waiting to get discovered. It doesn't anymore. AI search is very good at pointing people toward specific humans, and it isn't making that decision based on follower count. When Meltwater and LinkedIn analyzed 9.5 million AI citations this spring, LinkedIn came back as the second most-cited source on the internet, and 75% of those citations pointed at individual people rather than company pages.
Even more promising? More than half of the people getting cited had fewer than 10,000 followers.
What's standing between you and those citations usually isn't the quality of your expertise. It's that your expertise is scattered across the internet, vaguely described, and hard for a model to attach to one identifiable person.
So here's the afternoon that fixes it: pick the topic you want to own, rewrite your LinkedIn profile around it, publish one thing that's actually built to get cited, and connect the work you've already done so it reads as one expert instead of a dozen strangers.
Pick the one thing you want to be known for
Write down the topic you want to be the answer for, then keep narrowing until it feels a little uncomfortable.
"Marketing" is not a topic. "B2B marketing" is not a topic. Too broad. "Pricing and packaging for B2B SaaS companies at the Series A to Series B transition" is a topic, and it's specific enough that a model can attach to a human being.
You want to be the obvious answer to a question 500 people ask rather than the four hundredth plausible answer to a question 50,000 people ask. Niche is your superpower, baby. Two gut checks before you lock it in:
First, have you actually done the thing, not just read about it? That's the credential you can't manufacture and it's what makes your examples specific enough to be worth quoting.
Second, can you say something about it that your competitors can't, whether that's a number from your own work or a position you'd actually defend in an argument? If your take is the consensus take, there's no reason for a model to reach for you instead of anybody else.
Rewrite your LinkedIn profile around that one thing
Your personal LinkedIn page is one you fully control. Plus, it's the second most-cited domain in AI search, and three-quarters of those citations go to individual profiles rather than company pages.
Your headline should make a claim about a topic instead of describing a job. "Director of Product Marketing at Acme" tells a model where you work, which it does not need. "Pricing and packaging for Series A to B SaaS | Director of PMM, Acme" tells it what to cite you for, which is the entire point.
Your About section lives or dies in the first two lines, because that's what gets read before the fold and before anything truncates. Lead with what you know and how long you've been doing it, not your career narrative and definitely not your philosophy about helping teams unlock their potential. "Eleven years pricing B2B software; recently repriced a $40M ARR product line" does more work than three paragraphs of title jargon.
Your Featured section is where you gather everything. Every talk, article, podcast appearance, and mention you can dig up. More on why that matters in a minute.
Publish something that's built to get cited
When the Meltwater team broke down the structure of the most-cited LinkedIn articles, the pattern was almost uniform.
100% used bullets or numbered lists
92% used clear section headings
75% named specific companies, tools, or products
67% included hard numbers
50% used some kind of comparison or evaluation framework
By format, the winners were "best X" lists (54%), side-by-side comparisons (50%), and "how to choose" guides (33%). Pure opinion pieces barely registered.
None of that is a coincidence. It's a description of what's easy to lift into an answer. A model assembling a response needs pieces it can actually extract: a claim, a number, a named thing, a criterion for deciding. Give it a theory piece that swirls around a topic without landing anywhere, and there's nothing to grab onto.
Take the thing you already know cold and write it in that shape.
"How to choose a pricing consultant."
"Usage-based versus seat-based pricing, and the point where each one breaks."
"The four pricing mistakes I watch Series A companies make every single time."
If you want to REALLY nail this, create a named framework, as this sticky language is just one more lever you can pull toward becoming a named expert.
Two more findings worth designing around: 72% of cited content was original rather than reshared, and 48% of it had been published within the previous three months. Recency compounds here, which means one real post a week beats a burst of ten posts followed by a quarter of silence. Also worth knowing, text posts made up 72% of cited content and video only 11%, so you do not need production value. You need words with structure.
Connect the trail you've already left behind
Here's the problem nobody warns you about: your name appearing on 200 pages does not tell a model that those 200 pages are the same person.
You quoted in a trade publication, you on somebody's podcast, you with three years of good posts, you on a panel in 2023? Those are all separate floating strings until something connects them.
Most people spend a decade scattering evidence of their expertise and never once tie it together.
Pick one exact version of your name and use it everywhere from now on. "Kate Smith" on one profile and "Katherine Smith" on another and "Kate Smith, MBA" on a third reads as three different people. Same goes for how you phrase your topic. (I had to go back and update alllll my old bios across the internet when I made this pivot, and it was genuinely tedious, but it's the thing that makes everything else work.)
Claim the bylines that already exist. If work you wrote went out under "Team" or the company account, ask for it to be reassigned to your name with a one-line bio at the bottom. That's ten minutes of somebody else's time and zero new writing from you, which makes it the best trade available.
Take the author page if your company will give you one. It's a strong second anchor. If they won't, don't go build a separate personal website to compensate, because two half-maintained homes are worse than one good profile.
Then send one thing outward before you close the laptop. A podcast pitch with an actual proof point attached, not "I'd love to come on the show." A reply to a journalist request on LinkedIn or Qwoted. A guest post pitch to a publication that already ranks for your topic. One pitch, today, using the exact name and title you just standardized so it stacks instead of scattering.
What you can skip
Chasing followers, first of all. In the Meltwater data, 51% of cited creators had fewer than 10,000 followers, and in a parallel Semrush study, most of the cited posts had somewhere between 15 and 25 reactions.
Nobody's algorithm was blessing these people. The engine just found the clearest available answer and used it.
And you can skip posting daily. Consistency matters; volume theater doesn't. The LinkedIn article-writers getting cited cluster around 3-4 posts per week, not a few per day.
How to tell if it's working
Open ChatGPT, Perplexity, and Google's AI Mode and ask all three the same questions: "Who are the leading experts on [your topic]?", "Who should I follow to learn about [your topic]?", and then just your name, followed by your name plus your company. Screenshot all of it.
That last one is the prompt that matters most for your career, because it's what a recruiter or a prospective client is actually going to type. It's also the one that moves first.
Watch for the description going from wrong or empty to accurate, which usually lands inside thirty days. After that, you'll start showing up in the topic answers, and eventually your profile becomes a source the engines keep coming back to.
Rerun everything at thirty and sixty days, and remember that these answers are probabilistic. One flat rerun is weather, not a verdict. Track it in a spreadsheet with four columns (prompt, engine, date, named or not) and watch the pattern instead of the one-off.
Get clear and get cited
You're a person, which makes you nameable by default.
What's been missing is a version of you that's easy to find, consistent enough to recognize, and specific enough to be worth quoting, and that's genuinely one afternoon of work that almost nobody in your field has bothered to do.
Need help with this? I'm offering one-on-one consulting sessions for people who need personalized recommendations on how to jump-start this work. Use the form below to get in touch.
