The question I hear most from marketing teams right now is some version of: our traffic looks different than it did two years ago, and we do not understand why. AI search is a significant part of the answer, and this talk is built around explaining that shift clearly and giving people something to do with the information.

What Has Actually Changed

Search has been running on the same basic model for twenty-five years: user types query, engine returns ranked list of links. That model is not gone, but it is being layered under and around by something different. Google AI Overviews answer questions before the user reaches the organic results. ChatGPT and Perplexity answer queries without sending traffic to a source at all, or send it to one citation rather than ten.

The traffic impact is uneven. Informational queries — the ones that used to drive top-of-funnel content — are losing click volume fastest. Transactional and navigational queries still drive traditional search behavior. Understanding which parts of your traffic are at risk and which are not is the first thing this talk helps an audience figure out.

There is also a quality-of-traffic shift that is less obvious than the volume change. The clicks that AI search does pass through tend to go to pages that appear authoritative to a language model — which means clear entity signals, structured data, content that answers the question directly rather than working up to it slowly. So the sites that adapt well are seeing their click quality hold even as volume shifts. The sites that do not adapt are losing both.

Session Structure

I organize this talk into four parts:

  • What AI search engines are and how they work — not a deep technical primer, but enough model literacy to understand why your content might or might not be cited
  • The traffic signal shift — what the data tends to look like when AI search is pulling clicks away from traditional results, and how to read your analytics accordingly
  • What earns visibility in AI responses — entity clarity, structured data, content authority signals, and the difference between being indexed and being trusted
  • Practical next steps — a short list of things a team can actually prioritize rather than a long wish list

The session runs 45 to 60 minutes for most conference or corporate settings, with 15 minutes of Q&A built in. I can extend it to a 90-minute version with a deeper workshop component for smaller, more technical audiences.

Reading the Analytics for AI Search Signals

One of the most useful things I can give a marketing team is a framework for reading their own Search Console data to understand what is happening. The patterns are recognizable once you know what to look for: informational queries where impressions held but clicks dropped, position changes that did not correspond to ranking shifts, and specific query clusters where AI Overview appearance is suppressing CTR even on strong-ranking pages.

Google Search Console now shows AI Overview appearances for some accounts. For accounts where that data is not yet available, you can proxy it by looking at queries where your position held but CTR dropped by more than 30 percent over the past six to twelve months compared to the same period prior. That pattern, adjusted for seasonality and algorithm changes, is often a signal that an AI Overview is intercepting clicks above your organic result.

Why Schema Is Central to This

My specialty for the past several years has been advanced, custom schema markup — the structured data layer that helps machines understand what a page is about, who the author is, what entities are mentioned, and how the content relates to a knowledge graph. That work, done through Salterra Digital Services, gives me a concrete handle on one of the most actionable parts of AI search optimization.

Schema is not the whole answer to AI visibility, but it is one of the clearest technical levers a team has. I cover it practically in this session rather than abstractly. A page with well-implemented Author schema and Organization schema with entity links to authoritative external sources signals to a language model that the content has a verifiable source. A page without that structured context is harder for a model to confidently attribute and cite. That is a meaningful difference in how AI search treats the content.

Content Structure for AI Readability

Beyond schema, there are content patterns that correlate with AI citation. Direct answers early in the content, clear definition of the topic entity in the first paragraph, FAQ structures that match the natural language queries people actually ask, and content that does not bury the answer under context-building paragraphs. These are not just good writing practices — they are legible to a language model in a way that rewards the content with citation.

I cover a practical content audit framework in this session that a team can apply to their existing library. The question is not whether to rewrite everything — it is which pages are closest to being AI-citable and what small structural changes would get them there.

The Counter-Case: What AI Search Does Not Change

A question I get from every audience is some version of: does this mean our existing SEO work is wasted? The honest answer is no, but the emphasis shifts. Traditional ranking signals — technical health, inbound links, content relevance — still matter for traditional organic results, which still drive a significant share of search-referred traffic. The sites losing the most to AI search are the ones that were almost entirely dependent on informational top-of-funnel traffic from queries that AI Overviews now absorb completely.

Sites with strong transactional intent queries, local search signals, or navigational traffic are seeing much smaller AI search impacts. The session is honest about that distribution rather than treating AI search as a catastrophic wipeout of all organic traffic — because that framing is not accurate for most businesses.

Who This Session Is For

This talk works well for audiences that range from marketing managers to experienced SEO practitioners. I calibrate depth based on what the organizer tells me about the room. A corporate marketing team that has been doing traditional SEO for a few years is a very different audience from a conference full of agency professionals who live in this daily.

For a more focused session specifically on generative engine optimization tactics, the GEO speaker page describes a narrower, more technical version of this conversation.

Teaching This Regularly

I run live cohorts through SEO University, and AI search is now a core unit in that curriculum — not a supplemental topic. Teaching it in a live format, with students asking real questions, keeps the material grounded and current. What works in the classroom tends to work in a conference session for the same reason: it has been tested against people who will push back if something does not make sense.

More on formats, availability, and how to bring this session to your event is on the main SEO speaker page.