The Recommendation That’s Two Years Late
Imagine for a moment: an event planner asks AI, ChatGPT, Google, or Claude, for a keynote speaker on resilience with a $5,000 budget. It comes back with three names: one retired last year, one pivoted topics entirely and one hasn’t been in that budget bracket for two years.
This is a real problem AI keeps running into, and it’s already affecting speakers like you. Here’s the freshness problem in one line: AI pulled from whatever data it could find, at whatever point it was trained, decided that data was reliable, and presented it. Nothing since then corrected the record: and that’s a PROBLEM.
Why AI Still Has Your Old Reel
AI search tools work by retrieving sources first, then answering from whatever they retrieved. When there are several sources saying roughly the same thing about you, the newer ones tend to win. One 2026 analysis of AI citation patterns found that about half of the content these systems cite is less than thirteen weeks old, and that a page can lose something like half its citation potential within a year of publishing. That’s not a universal law, and results vary by tool and topic. But, it points at something speakers can feel intuitively and rarely act on: freshness isn’t a ranking nice-to-have here.
Maybe you’re thinking, “my site is current.” Great, that’s one real step. Same if your reel is recent. But AI can still be citing a bureau bio or a podcast appearance from three years ago right alongside it, no longer current, because that source never got refreshed and nothing newer showed up to outweigh it. I’ve seen this happen with clients: the old identity or inconsistency in how you present yourself can become an anchor to what you’re trying to do today.
Being Recommended Once Isn’t the Same as Staying Recommended
Known, Recommended, Chosen is the framework I use for this: identity (who you are across the web), third-party citations (PR, Reddit, LinkedIn, wherever else you’re mentioned), and ultimately, in Chosen, whether you’re speaking to what your CUSTOMER and prospects are actually thinking, wanting, and typing. “Recommended” isn’t a badge you earn once. It has to be renewed, the same way Known and Chosen do.
That’s a harder thing to hear than “you’re invisible to AI,” because invisible has a clean fix: get found. Outdated is messier. You were found. The version of you that’s still circulating just isn’t the one you’d want in the room.
What Actually Refreshes the Signal
Not a full rebrand. A steady drip of current, consistent signal across the places AI tools actually pull from:
Bureau and directory listings. If you’re listed anywhere, that listing is a source. An outdated one is actively working against you, not just sitting there neutral. These are often high domain authority (trusted) domains, which can further multiply the issue. Yes, it’s work to do “clean up” on these, but better to be right than a listing working against you.
Recent, topic-consistent appearances. Podcasts, panels, guest posts, LinkedIn posts. Not volume for its own sake. Consistency between what you’re saying now and what you want to be known for now.
Your own site’s structured data. Schema, bio pages, and talk descriptions that reflect your current topics, not the ones you spoke on two years ago. This is the unglamorous part, and it’s usually the one you have the most control over.
A domain mention now and then. Being written about, cited, or quoted by someone else carries more weight than saying it about yourself, and it’s one of the faster ways to reset how recent your signal looks.
Consistent positioning across your other owned properties. This isn’t just your website. Your LinkedIn profile, your speaker one-sheet (if’s on the web in some format), your bureau bio, any press kit you control directly. If they’re telling AI a different story than your site is, that inconsistency is itself a signal, and not the one you want.
This Is a Booking Problem, Not Just a Search Problem
Serendipity isn’t a strategy, and neither is assuming last year’s bio is still doing its job. An event planner running a cold search doesn’t know which version of you they’re getting. They just book the name AI hands them, or they move to the next one on the list. If you get one shot at the listing, you want it to be correct, yes?
If you’ve never actually seen what AI tools currently say, whether it’s you or the old version of you still circulating, that’s the place to start. The AI Visibility Audit runs targeted queries across ChatGPT, Claude, Perplexity, and Gemini and comes back with a written report on exactly what’s showing up, stale signal included.
If you want the fuller picture of how AI search changed speaker discovery in the first place, start with AI Search for Speakers. And if AI simply isn’t surfacing you at all yet, rather than surfacing an old version of you, this is the more likely piece for that.
