Written by Jonathan Dunnett
5 minute read
In Brief AI speaker recommendations are quietly changing how event planners find keynote talent. This post walks you through a two-minute self-search so you can see exactly where you stand, and explains what the results actually mean for your speaking business.
AI speaker recommendations are reshaping how event planners find talent
Most speakers have no idea where they stand.
Not all event planners are using AI tools yet. But enough are that if you’re not showing up in those results, you’re already missing opportunities you don’t know exist.
Here’s the thing about invisible losses: they don’t show up anywhere. No rejection email. No “we went in a different direction.” Just silence, and a booking that went to someone else.
I want you to do something before you read any further.
How Event Planners Are Using AI to Find Keynote Speakers
Open ChatGPT or Perplexity. Type in something close to this:
“I’m planning a leadership conference in [your city] with a budget of around $20,000. Can you recommend a keynote speaker?”
Use your topic. Use a realistic budget. Make it as close to an actual booking scenario as you can.
Then hit enter.
See who comes back.
If your name is there, good. Keep reading anyway, because what’s around your name matters too.
If your name isn’t there, that’s not a glitch. That’s your current visibility in a channel that event planners are quietly starting to use, and it’s worth understanding why.
Why AI Speaker Search Is Different From Google
AI search isn’t replacing Google. It isn’t replacing bureau relationships or referrals, but it is becoming part of how some event planners start their search, and that share is growing.
The dynamic that makes this particularly important is this: AI tools don’t just return a list of names. They return names with context. They explain why they’re recommending someone. They describe the speaker’s topic, their audience, their positioning.
Which means two things:
- First, if you’re showing up, what is the AI actually saying about you? Is it accurate? Is it compelling? Does it reflect the speaker you are today, or some version of you that existed three years ago based on whatever content was easiest to find?
- Second, if you’re not showing up, the AI has either never encountered enough about you to form a recommendation, or what it has found is too scattered, too thin, or too inconsistent to surface with confidence.
Both of those are fixable. But you can’t fix what you haven’t seen.
What Speaker Roulette Is Revealing About AI Visibility
I’ve been running a series called Speaker Roulette (subscribe – it’s free!), where I generate real booking scenarios and run them through multiple AI tools to see who shows up. The results have been genuinely interesting.
Big names missing entirely. Solid speakers showing up on one tool but not another. And in some cases, speakers whose own digital footprint is working against them, where a bureau listing says one thing, a website says something else, and the AI can’t reconcile the two into a coherent recommendation.
I’ll be publishing deeper findings from that research soon. But the self-search above will tell you more about your own situation in two minutes than any report I could write.
How to Read Your AI Search Results
It’s not just about whether your name appears.
Pay attention to:
- Who is appearing, and why. What are the AI tools saying about those speakers? What signals are they picking up on
- What topics are being associated with the results. Is the AI interpreting the scenario the way you would? Are the speakers it’s recommending genuinely your competition, or is something off in how the scenario is being read?
- What the AI says if you ask a follow-up. Try: “Are there other speakers on this topic I should consider?” and see what happens. Sometimes names that didn’t appear in the first response surface in the second.
If you’re not sure what event planners are actually looking for, this post covers exactly that.
Why Referrals Alone Are No Longer Enough
Referrals have always been the lifeblood of the speaking business. They still are. But referrals are serendipitous by nature. Someone has to think of you, at the right moment, for the right opportunity. You have limited control over when and whether that happens.
AI search is different. It’s systematic. It responds to signals. And those signals, your website structure, your content, your consistency of identity across platforms, are things you can actually work on.
Serendipity isn’t a strategy. Building visibility that compounds is.
Jonathan Dunnett helps keynote speakers get booked on purpose. He is a Council of Competitive Intelligence Fellow and brings systematic, intelligence-driven frameworks to speaker business development. Serendipity isn’t a strategy.
Frequently Asked Questions
What is an LLM, and why should I care about it as a speaker?
LLM stands for Large Language Model. It’s the technology behind AI tools like ChatGPT, Perplexity, Claude, and others. When someone types a question into one of these tools, the LLM generates a response based on everything it has learned from text across the internet, including websites, articles, directories, and more.
As a keynote speaker, you care about this because event planners are starting to use these tools to search for speakers. When someone types “recommend a keynote speaker on leadership for a corporate conference in Chicago,” the LLM generates a list. Whether your name appears on that list depends on how well the AI understands who you are and what you do.
Isn’t this just SEO with a new name?
Isn’t this just SEO with a new name?
Not exactly. Traditional SEO is about ranking in Google search results. LLM visibility is about being understood and surfaced by AI tools, and the two don’t always overlap.
A 2025 study by Ahrefs, analyzing 15,000 prompts across four AI tools, found that only 12% of URLs cited by ChatGPT, Gemini, and Copilot appear in Google’s top 10 results for the same prompt. That means ranking well on Google doesn’t guarantee you’ll show up in AI recommendations, and some speakers who don’t rank highly on Google are showing up in AI results because their content is clear, consistent, and well-structured across the web.
Perplexity is the notable exception. Nearly 1 in 3 of its citations do come from Google’s top 10, which means it behaves more like a traditional search engine than the other tools do. The practical implication for speakers: different AI tools pull from different places, and your visibility on one doesn’t guarantee visibility on another. That’s exactly why foundational work across all your digital properties matters more than trying to game any single platform.
They are related disciplines, and good foundational work helps both. But they are not the same thing.
Doesn’t optimising for AI feel like gaming the system?
This is one of the most common concerns I hear, and it’s worth addressing directly.
Optimising for AI visibility isn’t about manipulation. It’s about clarity. LLMs struggle to recommend you when your identity is inconsistent, your website is poorly structured, or your expertise is buried under vague language. Fixing those things doesn’t game anything. It just makes it easier for the AI (and for humans, for that matter) to understand what you do and who you do it for.
Think of it this way: if you walked into a networking event and nobody could tell from talking to you what you actually speak about, that’s not the room’s fault. Clarity is always in your interest.
The other opportunity? It also helps you think about the customer journey: they have questions that you can answer, and this can help you be cited as you address that on owned and non-owned assets and properties across the web.
How do AI tools decide which speakers to recommend?
AI tools pull from a wide range of sources: your website, your LinkedIn profile, Reddit, your bureau listings, articles and interviews you’ve been featured in, podcast appearances, and more. They build a picture of who you are based on the consistency and clarity of that information across all of those sources.
A few things that influence whether you show up: how clearly your website describes your speaking topics, whether “keynote speaker” appears prominently and repeatedly across your digital properties, how consistent your positioning is across platforms, and how much credible third-party content exists about you.
There is no single algorithm to crack. Different tools weight sources differently, which is part of why running the self-search across more than one tool is worth doing.
I’ve been speaking for years and have a strong reputation. Why wouldn’t I already be showing up?
Reputation and visibility are not the same thing, at least not to an AI.
A strong reputation built through relationships, referrals, and word of mouth is genuinely valuable. But LLMs can only work with what they can find and parse online. If your website hasn’t been updated in a few years, if your positioning has shifted but your digital footprint hasn’t caught up, or if your expertise is described differently on your website than on your bureau listing, the AI gets a confused picture. A confused picture doesn’t lead to a confident recommendation.
This is actually one of the more surprising findings from the Speaker Roulette research: speakers with strong real-world reputations who are essentially invisible in AI search, not because they aren’t credible, but because their online presence hasn’t kept pace.
Is this just a passing trend, or is it worth investing time in?
That’s a fair question. AI search is still evolving, and nobody can say with certainty exactly what it will look like in three years.
What we can say is this: the direction of travel is clear. AI tools are becoming a more common starting point for research, including for event planners looking for speakers. Getting your foundational visibility right now, while most speakers aren’t paying attention to it, is a meaningful advantage. That window won’t stay open indefinitely.
Where do I start if I want to improve my AI visibility?
The self-search described in this post is the right first step. See where you stand before you try to fix anything.
After that, the foundational work involves three areas: your website structure (headings, page architecture, dedicated speaking page), your identity consistency across platforms (website, LinkedIn, bureau listings), and the clarity of your positioning (are you easy to understand in one sentence?).
If you want to work through all of this in a structured way, I’m running a small group workshop covering exactly these areas, with enough time to apply it to your specific situation. Ten spots only.