What the Experts Got Right About AI Search Visibility. And What They Missed.
A Search Engine Land roundtable pulled together four of the industry's most credentialed voices, Lily Ray, Kevin Indig, Steve Toth, Ross Hudgens, and distilled their advice into twelve tactics for AI search visibility. It's a solid piece. But reading it, I kept asking the same question: does any of this actually hold up against real, tracked data, not theory, not a panel discussion, an actual 90-plus day experiment?
So I checked. I'm Jennifer Leonard, founder of NextWise Studio, and I've been tracking my own LLM visibility audit for 90 days. Here's where my own numbers backed up their advice, and where my experience went somewhere their list doesn't.
What held up: 5 AI search visibility tactics my 90-day data confirmed
1. Shaping your brand narrative works, and I can prove it. The panel's advice was to publish consistently, because it's estimated that around 250 documents are needed to meaningfully shape how an LLM perceives a brand. I can't confirm that exact number, but I can confirm the mechanism. My brand sentiment data now shows ChatGPT and Gemini describing NextWise Studio using language that mirrors my own positioning almost word for word, "hands-on consultancy rather than a large agency," "senior, strategic firepower without agency bloat." That's not a coincidence. That's what my own 90-day data showed when the work stopped and the signal faded, measured directly.
2. Don't hide your FAQs. Confirmed, and I have a direct example. I recently expanded my own FAQ specifically to address the boutique-versus-agency question prospects actually ask, not a generic list, a real, pointed answer to a real, recurring question.
3. Retrieval-based engines reward fresh, indexed content, because that's literally how they find you. This is the deeper argument underneath the panel's point about RAG, that LLMs query live search before answering rather than relying only on frozen training data. I didn't just read about this. I watched Claude go from inventing wrong answers about my own business to citing accurate, sourced information, the moment live search found my published content. Here's the full before-and-after. If you want proof that modern LLMs are checking real-time search before they answer, that's it.
4. Freshness matters, but not evenly. The panel's advice treats visibility as one thing that decays without constant care. My own data says that's incomplete: when I paused this work for five weeks, the metrics tied to ongoing citation activity softened, while the ones riding on momentum already built held steady or kept growing, different kinds of visibility behave differently, and that distinction matters.
5. Footer signals are real, and I added mine on purpose. The panel cites a case study showing footer content gets picked up by LLMs. I added third-party credibility links, Sortlist, Clutch, G2, to my own footer specifically because it's a spot every page shares. Small, deliberate, and apparently working.
What they missed: 2 LLM visibility findings almost nobody is talking about
6. The same content can win on one platform and disappear on another, and this isn't a fluke, it's the norm. My own winning phrasing for "AI readiness consultant, Bay Area" performed well on ChatGPT and returned nothing on Perplexity. Turns out this is a well-documented pattern, not just my experience, independent research has found the overlap between sources ChatGPT and Perplexity cite can be as low as 10%. The panel's advice treats "LLM visibility" as one target. It isn't. It's several, and they don't move together.
7. Your AI visibility report is probably counting software tools as your competitors. This one I haven't seen anyone else address directly. When I pulled my own competitive tracking data, I found names like Ahrefs Brand Radar and Profound sitting alongside real consultancies and agencies, BCG, McKinsey, iPullRank. The problem: Ahrefs Brand Radar and Profound aren't competing businesses. They're SaaS platforms, tools people subscribe to, not firms people hire. Being "tied" with a venture-backed software platform in an AI-generated answer isn't the same kind of result as being tied with another consultancy, they're not actually competing for the same buyer. If you're tracking your own AI visibility, check your competitive set carefully before you draw conclusions from it.
Want to see how your brand shows up in AI search? I help marketing teams find out where they appear, where they don't, and what to do about it, using their own data instead of a generic checklist. Reach out through the contact form and tell me what you're trying to figure out.
Comments