In this solo presentation, Jason Berkowitz examines how AI and large language models (LLMs) are reshaping search engine optimization (SEO). He begins by citing data that Google still commands the overwhelming majority of search queries, debunking hype that other platforms are overtaking it. Berkowitz then highlights the inconsistency of LLM outputs (ChatGPT, Gemini, Perplexity), which makes brand awareness monitoring difficult and offers no proven framework for success. He explains that AI-generated overviews on Google boost impressions but dramatically reduce clicks, especially for top‑of‑the‑funnel informational queries, and that these overviews often contain hallucinations and will soon be ad‑laden. The core strategic advice is to pivot SEO investments toward middle‑ and bottom‑of‑funnel assets—service, product, and category pages, as well as thought‑leadership tied to revenue—rather than purely informational content. Berkowitz also warns that AI‑driven changes are corrupting historical performance data, making forecasting and seasonal projections unreliable. He concludes that the future SEO playbook will likely dissolve SEO’s siloed status, integrating more tightly with broader marketing initiatives, while still offering opportunities to acquire new customers and revenue in the current environment.
Chapters
So there's this 2-letter acronym that's making the rounds as it has been for a year now, AI. And one question I tend to get asked just about weekly from agency partners to clients to people in my network is AI's impact on SEO. So first, if you haven't seen our article on Break the Web about AI and SEO statistics, go check it out because it really helps cut through the noise about what people are chatting about online versus what the data shows. And TL;DR, the data shows that Google still holds the largest market share. They're not— no platform is even making a dent in the amount of Google search queries per day that are taking place versus any other platform. But what's crazy is there's also a lot of predictions and assumptions about what the future of SEO is really going to look like. SEO is dead. SEO is changing. SEO is evolving. And there's already companies out there, agencies that are advertising their SEO service, whether it's AEO or GEO or whatever acronym just ends up sticking, what brands need to be aware of is that nobody at this present time really has a true current proven framework of success. Nobody. LLMs like ChatGPT, Gemini, Perplexity, they're very consistent about giving inconsistent results, which makes trying to monitor brand awareness and brand visibility really, really challenging. I mean, with that, there are ways of course to audit and gauge a brand's perception on these platforms, but any proactive measures to increase a brand's visibility really might not be that different than what a really good SEO strategy would do. Increasing your brand presence, your brand relevance, your thought leadership, your brand awareness externally with good digital PR. Really good SEO already is helping visibility and brand awareness in these platforms, especially and most predominantly in the B2B space. And that might be just people who are using these type of platforms to make searches are are just more aware about tech to begin with. So when it comes to SEO specifically, yeah, the elephant in the room when you make a Google search sometimes is the AI overview. And in theory, the purpose of an AI overview sounds cool, but everyone's seen the data. Everyone in every industry, every brand is seeing clicks go down. They're seeing impressions go up if they're, especially if they're cited in an AI overview, But clicks are drastically going down. And the common consensus among various different people, consumers, everyday people on Reddit, is that AI overviews are still very inaccurate. They hallucinate. And we know there's a good chance they're probably going to be flooded with ads very soon. I mean, Google does have to make their money back from everything they're doing. The question is, what does this mean for user adoption? But, and I think that's yet to be determined how people are going to view AI as a search mechanism. AI overviews are most predominantly featured in informational, super top-of-the-funnel queries, really people who are just looking for a straight answer to a question. But the question we need to ask as brands is that lost traffic or lost clicks really and truly lost revenue? And sometimes topics that we have on our site that are losing traffic, might be so top of the funnel. It might be so informational that if somebody was clicking on your search result listing anyways, they probably weren't going to convert into revenue or sign up to an email newsletter or click onto your product page. So the question we need to be asking our SEO agencies really is about the bottom line. How are clicks and performance metrics to various different pages that bring in new users to the funnel and drive revenue impacting the bottom line. Again, SEO should be adapted to maybe more middle or bottom of the funnel. People that are looking for the right place to spend their money, where informational top of the funnel, people don't have their wallet in hand. They may not be ready to make a purchase, but focusing more again on middle of the funnel, even with thought leadership content, which is still very much possible. And then of course, resources being put into bottom of the funnel, service pages, product pages, category pages still can actively bring in revenue today. And if all that wasn't enough, you have Google's AI mode, which people still have the option to opt out of if they'd like in mobile and desktop searches. And AI mode already contains ads. So is this going to come full circle in a way and bring us to a place that we're already extremely familiar with, which is ads, then value, kind of like traditional search results, ads, then search, organic search result listings. So there's where we're seeing some of the biggest impacts on SEO and even more where marketing teams seem to be running into big roadblocks is forecasting and projections. At this moment, historical data is ruined, which means forecasting is now ruined. We've historically have helped countless clients with forecasting data and projections, especially for their leadership team. But now with all these different big shifts taking place, historical data is just kind of wiped out. And really good historical data for these type of forecasts is ideally 25 months. worth of data to account for seasonalities and such. So to wrap all this up, really, what is the future of the SEO playbook? Who knows? There's a good chance SEO will expand and SEO activities will expand beyond just SEO. And maybe thankfully SEO will stop being in its own little segmented silo in the corner of the marketing team because we're in a state of evolution. And SEO can help impact other marketing initiatives and other marketing initiatives can help impact SEO. And very much today, even with this current state of evolution that we're in right now in SEO, there still is the chance to bring in brand new raving fans. Even more importantly, bring in brand new revenue at this present moment.
