A talk exists for forty-five minutes. A structured record of that talk exists forever, and only the record is visible to the engines deciding whether to shortlist you. Build one speaker page that lists every talk with event, date, format and a verifying link, add Person and Event markup so machines can attribute it all to one entity, and push for an official listing at every future event you accept.
Speakers love to say the room was electric. Machines have never been in a room. If your decade on stage lives only in memories, photos and a bullet on your LinkedIn profile, then as far as any AI engine is concerned, it barely happened.
Why your speaking history is invisible to machines
Speaking is among the strongest credibility signals a professional can earn: an independent organization judged you worth an audience's time and put its name behind yours. The problem is purely one of evidence handling. Talks are ephemeral by nature, and the traces they leave are scattered and fragile: an agenda page that gets deleted when the conference site resets for next year, a YouTube upload titled "Day 2 Afternoon Session," a group photo, a tag in someone's post.
An engine trying to answer "who are the credible speakers on supply chain risk" cannot weigh what it cannot retrieve and attribute. It needs three things: a claim (you spoke), a verification (an independent page confirming it), and an attribution (certainty that the speaker named is the same entity as you). Most speaking histories fail on all three. The claim is vague, the verification link rotted years ago, and the attribution is muddied because the conference spelled your name differently than your website does.
This matters more now than it did in the ten-blue-links era, because answer engines return names, not result pages, and a growing slice of discovery ends inside the answer itself: zero-click searches grew from 56 percent to 69 percent in the year after AI Overviews launched, per SEO Sherpa's AI search statistics. When the answer is the whole product, being retrievable and verifiable is the whole game.
What should a speaker page include?
One page on your own site should be the canonical, permanent record of your speaking life. Event sites decay; your page does not. Here is the full checklist:
- Your canonical identity block: full name, one-sentence positioning, your standard 40 to 80 word bio and your one canonical headshot, identical to everywhere else you appear.
- Speaking topics phrased as problems: not "keynotes on innovation" but the three to five specific questions you address from stage, in the language event organizers and buyers actually use.
- The talk ledger: a dated, reverse-chronological list of talks. Each entry carries the talk title, the event name, the city or platform, the year, the format (keynote, panel, workshop), and links to the official listing and the recording where they exist.
- Proof artifacts: embedded or linked recordings, slides where shareable, and one or two organizer quotes with the person's name and role.
- Booking information: what you speak about, formats you accept and how to reach you. A speaker page that cannot be acted on wastes its own traffic.
- Structured data: Person markup for you, Event markup for the appearances, covered in the next section.
Every talk entry should let a stranger, or a machine, answer four questions without leaving the line: what did you present, where, when, and can it be verified. An entry that fails the fourth question is a claim, not a signal.
How do you structure the markup?
The visible page serves humans and retrieval. The markup underneath serves attribution, which is the step most speaker pages skip entirely. Two vocabulary types do the work: Person declares who you are and connects your profiles, and Event describes each appearance with you attached as performer. Google's structured data documentation explains the general mechanics; what follows is the minimal pattern applied to a talk entry.
The visible HTML for one entry stays simple and readable:
<article class="talk"> <h3>Pricing Under Uncertainty</h3> <p>Keynote, SaaS Growth Summit 2025, Austin, TX</p> <a href="https://example.com/summit-2025/agenda">Official agenda listing</a> <a href="https://example.com/watch/pricing-keynote">Recording, 42 min</a> </article>
And the JSON-LD that makes it attributable, placed once in the page head with one Event object per talk:
{
"@context": "https://schema.org",
"@type": "Event",
"name": "SaaS Growth Summit 2025",
"startDate": "2025-09-18",
"location": { "@type": "Place", "name": "Austin, TX" },
"performer": {
"@type": "Person",
"name": "Alex Rivera",
"url": "https://alexrivera.com",
"jobTitle": "Pricing Strategist"
}
}
Notice what the performer block accomplishes: it binds the appearance to your exact name and your canonical URL, so this talk compounds onto the same entity as your articles, your podcast episodes and your reviews. Without it, "A. Rivera" on a conference site and "Alex Rivera" on your homepage may never be confidently merged. The markup is not a trick and it earns nothing on its own; it is plumbing that stops your earned credibility leaking away.
The human job and the machine job of each element
A good speaker page does double duty on every element. Keep this split in mind while building:
| Element | Job for the human reader | Job for the machine |
|---|---|---|
| Identity block | Instant orientation: who is this | Anchor facts to merge with other mentions |
| Topics as problems | Lets an organizer imagine the session | Co-locates your name with buyer-phrased queries |
| Talk ledger with links | Demonstrates range and recency | Verifiable claims with independent corroboration |
| Recordings and quotes | Proof of stage quality before booking | Rich media and third-party language about you |
| Person and Event markup | Invisible | Unambiguous attribution of every appearance to you |
Make third parties confirm it
Your page is the claim; independent pages are the verification. From now on, treat the paper trail as part of every speaking negotiation. Before accepting a slot, ask the organizer three things: will there be a speaker listing on the event site with my name, title and site link; will the session be recorded and published; and may I have a quote from the program chair afterward. Organizers agree to these routinely when asked early, and almost never think of them unasked.
After the event, harvest within a week, exactly as you would after a podcast episode, a discipline laid out in the podcast guesting playbook. Archive the agenda page, since conference sites are notorious for wiping history each cycle. Add the entry to your ledger with its links. If the event published a recap that names you, link that too. Speakers who run this loop for a year build a verification trail that a machine can walk end to end, which is precisely what separates the shortlisted from the merely experienced, a divide we examine further in PEO for keynote speakers.
Where this fits in the bigger build
A machine-readable speaking history is one shelf of a larger evidence structure. It slots into the earned-proof layer of your proof portfolio, and its verified, third-party nature is what lets it punch far above owned content in an engine's weighing. If you are sequencing a full identity build, do the consolidation and markup work in the same pass; the first 90 days plan shows where the speaker page lands in the order of operations. And if you would rather hand over the construction, the audit, the markup and the organizer playbook are standard parts of our services.
Questions
What should a speaker page include? +
Does schema markup on a speaker page actually matter? +
My old talks have no trace online. Are they wasted? +
See what AI says about you today.
Start with a reading. We show you the words the engines return about your name, then map the fastest signal to move.
Get named →