Subscribe
Share
Share
Embed
Zero Click Marketing is a marketing strategy podcast about content marketing, audience research, and how brands grow when clicks matter less. Hosted by Amanda Natividad, Chief Evangelist at SparkToro, the show explores how marketers reach audiences, build influence, and earn attention in a zero-click internet. New to the show? Start with Episode 2: What Zero Click Marketing Actually Is.
Amanda Natividad [00:00:00]:
AI is changing SEO. I know. I'm sorry. I also wish we could go one fiscal quarter without someone inventing a new acronym and then acting like it descended from the heavens wearing a Patagonia vest. But unfortunately, this one is mattering more and more. The acronym is GEO or generative engine optimization. And depending on who you ask, GEO is either the future of marketing, a rebrand of SEO, a technical discipline, a content strategy, a consulting upsell, or a really effective way to make marketers feel like they're behind on yet another thing. And no, this is definitely not another phrase I coined. So today we're going to talk about what GEO actually is, what people are overclaiming, and what marketers should do about it. And here's my argument. GEO is what happens when search becomes zero click because if AI answers your buyer's question before they click your website, the game changes.
Amanda Natividad [00:01:01]:
Not completely, not in some SEO was dead, burn the blog, fire the content team way. No. But the game does change because now the question is not only can we rank, the questions become more and more broad. Can we be retrieved? Can we be understood? Can we be cited? Can we be remembered? Can we become part of the answer before someone ever visits our site? That is a zero click problem and now it is also the AI search problem. Google's own documentation says AI features like AI overviews, AI mode are now part of search and that site owners should still focus on helpful, reliable people first content, which is both true and frankly a little annoying because it sounds like the corporate version of just be yourself. Helpful, reliable, people first, great, love that for us. Also, can we get a little more specific? That's what this episode is for.
Amanda Natividad [00:02:02]:
I'm Amanda Nativadad. Welcome to Zero Click Marketing.
Amanda Natividad [00:02:08]:
Ahrefs' December 2025 analysis found that AI overviews are now associated with a 58% lower average click-through rate for the top ranking page. In their dataset, position one click-through rate for informational keywords dropped from 7.6% in December 2023 to 3.9% in December 2025. But for keywords that triggered AI overviews, position one click-through rate dropped from 7.3% to just 1.6%. So no, AI overviews don't eliminate all clicks, but they do appear to siphon away a meaningful share of the clicks top ranking pages used to earn and that's the important part. AI search does not have to destroy your traffic to change our marketing strategy. It only has to change where influence happens and increasingly influence happens before the click. That's the whole zero click marketing thesis. So let's talk about GEO. The term GEO comes from the paper GEO, generative engine optimization, which defines it as a framework for improving visibility in generative engine responses.
Amanda Natividad [00:03:31]:
The researchers found that GEO methods can improve visibility, but that what works varies by domain, which is a very academic way of saying there is no universal AI citation hack. It's frustrating because we marketers love certainty. We love a checklist. We love being told there are four steps, seven secrets, or one counterintuitive trick. But GEO is not one weird trick. It's a set of practices that make your content easier for AI mediated systems and most importantly, humans to find, understand, trust, summarize, and cite. That's it. And if that sounds kind of boring, good. Boring is often where the money is. So I want to break this down into four mechanics, not four hacks, four mechanics. Number one, retrievability. Number two, extractability. Number three, credibility. Number four, public evidence. Let's start with retrievability. Retrievability means can the system access a darn thing in the first place?
Amanda Natividad [00:04:45]:
This is the unsexy technical layer. Before your content can be cited, summarized, or included in an AI answer, it has to be discoverable. It has to be crawlable. It has to be renderable. It has to be reachable through whatever retrieval system the AI tool is using. This is where marketers may need to work with developers, SEOs, or technical teams. You want to audit your robots.txt file, not because robots text guarantees inclusion, it doesn't, but because you do not want to accidentally block the crawlers you actually care about. AI documents separate crawlers including GPT bot and OAI search bot and says site owners can use robots text tags to manage how their sites and content work with OpenAI's AI systems. Perplexity also documents PerplexityBot and says webmasters can use robots.txt tags to manage how other sites interact with perplexity. But this is not a perfect little robot utopia.
Amanda Natividad [00:05:53]:
Perplexity says Perplexity bot respects robot's text while Cloudflare has publicly accused perplexity of using undeclared crawlers to evade no crawl directives. So the practical advice is not allow every AI bot. It's know what you are blocking, know what you are allowing and make those choices intentionally. But the practical takeaway is simple. Audit your robots.txt file. Make sure you are not unintentionally blocking the systems you want access from. Also, know what each crawler is for. Google-Extended, for example. It's not the same thing as ordinary Googlebot. Google describes Google-Extended as a control that lets sites manage whether their content helps improve Gemini apps and related APIs, not as a normal crawler for showing up in search. So please do not go into Slack and say, "Amanda told us to allow all the AI bots or we'll advantage from the internet." Amanda did not say that.
Amanda Natividad [00:07:02]:
Amanda said, "Know what you are blocking, know what you are allowing and make those decisions intentionally. And you know what? Maybe if you don't know what to do here, ask my friend Britney Muller or Mike King at iPullRank or Lily Ray at Algorithmic. They know a lot more about all of these technical implications than I do. Another piece of retrievability is renderability. If your most important answer is buried behind JavaScript, tabs, accordions, or app-like interfaces, you might have a visibility problem. Google's JavaScript SEO guidance says site owners should make sure Google can access, render, and index JavaScript powered content properly. And again, I'm using Google as a baseline here because Google's documentation is public and mature, but the broader principle applies. Don't make machines solve a puzzle just to understand your best ideas. Also, keep your XML site maps clean. SiteMaps help search engines discover pages and understand what they change.
Amanda Natividad [00:08:11]:
Google's site map documentation recommends using site maps to tell search engines about pages, images, videos, and other files on your site. I would be cautious about calling these AI specific site maps because there's not a universal AI site map standard that every engine has agreed to, but clean site maps still matter. Prioritize your most useful indexable answer rich pages. Not every single tag page or that dusty little landing page from 2019 that exists because someone once ran a campaign called Revenue Palooza. Your best stuff. The pages that explain your category answer real questions, include original evidence and that would genuinely help someone make a decision. There is also this emerging idea of an llms.txt file. The llms.txt proposal from Jeremy Howard suggests adding a markdown file at the root of your website to provide information that helps LLMs use your site at inference time.
Amanda Natividad [00:09:16]:
The proposal describes it as a way to offer brief background information, guidance, and links to useful markdown files or other resources. I like this idea. It's lightweight. It's plain language. It forces you to say, "Here are the most useful things on our site." That alone is probably valuable, but I wouldn't oversell it. Don't think we deployed LLMs text, therefore a Claude will rank us first. No and not just because AI ranking isn't real. Instead, keep this in mind. This is an emerging convention. It may help. It's low risk. It also forces useful internal clarity. That is an example of retrievability. Can systems find and access your content? Now, let's talk about extractability. Extractability means once a system finds your content, can it easily understand and reuse the useful parts? This is where GEO starts to overlap with good writing and that's inconvenient because everyone wants a technical hack and sometimes the answer is write a better paragraph.
Amanda Natividad [00:10:26]:
For AI search, your content needs to be easy to parse. Clear definitions, specific claims, useful headings, concrete examples, short answer blocks, evidence close to the claim, comparison language, original frameworks. So if someone asks, "How is GEO different from SEO?" And your answer is buried halfway down the page under a heading called The Evolution of Digital Discovery. Please stop that, say the thing, then explain the thing, then prove the thing. This is where I think marketers can learn something from journalism, documentation and product marketing. A good answer rich page does not make people work too hard and increasingly I think content teams should treat more of their content like reference assets. Not every piece of content needs a grand narrative arc, although I mean some of it should, right? It does give people a reason to care. It's just that some content should be designed to become the canonical explanation of an idea.
Amanda Natividad [00:11:27]:
A buyer should be able to read it. A salesperson should be able to quote it. A journalist should be able to cite it. An AI system should be able to summarize it without mangling it beyond recognition. That is extractability. And by the way, this does not mean write robotic content. It does not mean stripping out voice. It means writing your best ideas as standalone statements, clear quotable lines that can survive outside the paragraph they came from. A strong point of view is more extractable when it is named, structured, and supported. For example, GEO is zero click SEO. That's a standalone statement. Traffic was never the whole point. That's also a standalone statement. Your marketing needs more standalone statements because if your idea is floating in a 2000 word fog bank, neither humans nor machines are going to know what to do with it. The third mechanic is credibility.
Amanda Natividad [00:12:29]:
This is the part where the internet's old sins come back to haunt us because for years a lot of SEO content was written to look authoritative without actually being authoritative. You know the genre. What is customer research a complete guide written by someone who has never spoken to a customer, never conducted research and appears to have learned about the topic by reading eight other articles called What is Customer Research A Complete Guide. It is the lasagna that nobody wants to eat but cooks keep serving. Layer upon layer of rephrased sameness and AI answers make this problem weirder because sometimes it summarizes the consensus, but sometimes it pulls in the thing that stands out and sometimes the thing that stands out is not the thing you wish it had found. That's why credibility matters. It gives you a stronger public record than generic content, stale claims, and whatever random artifact happens to be most distinctive.
Amanda Natividad [00:13:36]:
Credibility means your content gives people and search and AI tools a reason to trust this answer from you because if everyone publishes the same generic answer, why should an AI system cite you? Why should a buyer trust you? Why should anyone remember? You. Credibility means your content has reasons to be trusted. That can include original research, first party data, expert quotes, clear methodology, customer examples, specific experience, screenshots, name sources, a real point of view, author expertise, transparent limitations. The original GEO research tested optimization methods like adding citations, quotations and statistics, and it found visibility improvements, though again, with variation across domains. This is not just an AI thing. This is a trust thing. You probably have proprietary knowledge sitting around your company right now. Sales calls, support tickets, customer interviews, usage data, podcast conversations, internal experts, competitive patterns. That's the good stuff. That is the material that can make your content meaningfully different from the commodity answers already clogging up the web, which brings us to the fourth mechanic, public evidence.
Amanda Natividad [00:15:01]:
This is the most zero click part of GEO. AI systems and people learn from all the discovery destinations online. Reddit threads, YouTube videos, LinkedIn posts, reviews, comparison pages, third party mentions, public conversations, and many more. Unfortunately, those are all the places you don't own, but that's what makes them more credible to your audience. So the public record around your brand is thin, confusing, outdated, or non-existent. You have a problem. I mean public record very literally. What does the internet seem to know about you? What do customers say? What comparisons exist? What questions come up repeatedly? What misconceptions go uncorrected? What phrases are associated with your brand? What proof exists outside your own website? This is why zero click marketing matters so much because in a zero click world, your brand is often being evaluated in places you don't control. And in an AI search world, those places can become inputs into someone else's answer.
Amanda Natividad [00:16:11]:
This is where marketers need to stop thinking only in terms of owned content and start thinking in terms of distributed evidence. Your website matters, yes, but so does the conversation around your website. So does the trail of proof and whether your category explanation exists in enough credible places that both humans and machines can connect the dots. This is also why I don't like the phrase AI optimization because it sounds too narrow because the goal is not to trick the robots. The goal is to become easier to trust by people first robots second. So what do you actually do this week? First, pick one page. One. Choose a page that matters. Ideally, a page that answers a question your buyers actually ask. Maybe it's your comparison page. Maybe it's a high intent blog post. Maybe it's your best research report. Then ask five questions. Question one.
Amanda Natividad [00:17:13]:
Can this page be found and rendered? Check robots text. Check indexing. Check whether the important content is visible. Check whether it relies on fragile JavaScript. Check whether it appears in your site map. Question two. Does this page answer the core question clearly and early? Not eventually early. Question three. Are the claims extractable? Do you have clear headlines, definitions, examples, short answer sections, summary bullets? Question four. Are the claims supported? Do you include sources, data, quotes, or experience? Question five. Does this page connect to the broader public record? Can people find related proof elsewhere? Are there customer stories, videos, podcast appearances, third party mentions, community discussions, social posts? Then improve the page, not by stuffing keywords, by making it more useful. Add the definition, example, caveat, comparison, source. Add the original point of view and what to do next and then crucially repurpose the idea offsite.
Amanda Natividad [00:18:30]:
Turn the best answer into a LinkedIn post. Talk about it on a podcast. Pitch it as a guest contribution. Use it in a webinar. Put it in your newsletter. Answer the question in the places where your audience already hangs out because GEO is not just a website project. It's a visibility project that requires you to build trust and strengthen your public record. That's the real reason I think marketers should care about GEO, but not panic about GEO because the people who panic will chase tactics, ones that are probably going to change in the next major LLM update. People who care about GEO ask questions like, "How do I become one of the most useful, trusted, and citable sources in our category?" That's much harder to answer, but the work that supports that is a lot more durable and it's not that different from what good marketing has always been.
Amanda Natividad [00:19:28]:
Be findable, be usable, be credible, be memorable. Show up where your audience learns. Say something worth repeating. Create evidence that outlives the click. That's it. That's the game. GEO is not magic. It's a reminder that when search becomes an answer engine, your content has to do a lot more than attract traffic. It has to earn its place in the answer. And if that sounds familiar, it should because that's zero click marketing. And if you enjoy this episode, why have I got the book for you? The new book, Zero Click Marketing by Rand Fishkin and yours truly is available for pre-order now. Go to zeroclickmarketing.co/book to reserve your copy today. Pre-ordering direct like this ensures that you get discounts off SparkToro and AlertMouse as well as access to exclusive webinars that Rand and I are running this summer. Plus you'll be supporting an indie publisher, which is what cool kids do.
Amanda Natividad [00:20:35]:
All right. See you next week, cool kids.