Table of contents

Chapter 1: The shift from Google to ChatGPT

For twenty-five years, the playbook was simple: rank on Google, get traffic, make sales. Every business owner with a website understood the basic transaction. You put content on the internet, Google indexed it, and people clicked through from search results to your pages. That era is ending faster than most people realize.

Something different is happening now. When someone types "best accounting software for a small restaurant" into ChatGPT, they don't get ten blue links. They get a direct answer. A recommendation. Often with a specific product name, a reason to choose it, and sometimes a link. That answer might send a buyer straight to one business and completely skip the other nine that would have shown up on a Google results page. The question this entire book tries to answer is: how do you become the business ChatGPT recommends?

How ChatGPT search works vs. traditional search

Google works by crawling billions of web pages, indexing them, and ranking them by hundreds of signals (backlinks, page speed, keyword relevance, domain authority, and so on). When you search, you get a ranked list and you choose which result to click. The website owner's job is to appear as high as possible in that list.

ChatGPT works differently at every step. It was trained on a massive text dataset, so it already "knows" a lot about brands, products, and industries from what it absorbed during training. When you ask it a question, it doesn't show you a list of websites. It synthesizes an answer by combining its training knowledge with real-time web search results (when browsing is enabled). It then presents a single, conversational response that often names specific brands, products, or services.

The shift matters because user behavior has changed with it. A Google search is a starting point for research. A ChatGPT answer feels like a conclusion. When ChatGPT says "for a small restaurant, consider FreshBooks because it handles tip tracking and integrates with most POS systems," many users treat that as a trusted recommendation and go straight to FreshBooks. No comparison shopping. No clicking through five blog posts. Just a direct path from question to purchase.

The numbers behind the shift

ChatGPT now has over 883 million monthly active users. To put that in perspective, it took Google about eight years to reach that number. ChatGPT did it in under three. Meanwhile, Gartner's research predicts traditional search engine volume will drop 25% by the end of 2026, driven by people switching to AI chatbots and virtual assistants for their information needs.

Google itself has noticed. Its AI Overviews (the AI-generated summaries at the top of search results) now appear in roughly 55% of all Google searches. That means even if you're still focused on Google, the game has already changed there too. The top of the results page is no longer a list of links you optimized for. It's an AI-generated answer that may or may not mention your business.

Here's what those numbers mean in practical terms: nearly a billion people are now asking an AI for recommendations. Most of them used to type those same questions into Google. And when ChatGPT answers, it mentions brands only about 20% of the time, which means the brands that do get mentioned are capturing a wildly disproportionate share of attention.

Clearing up the jargon: AEO, GEO, AI SEO, and the rest

Before we go further, let's sort out the terminology. This space has exploded with acronyms over the past two years, and people use them inconsistently. Here's what each one actually means and how they relate to each other:

The short version: AEO, GEO, LLMO, and "AI SEO" (in its second meaning) all describe the same core activity. They're all about making your business the one that AI recommends. This book uses "AEO" as the default term because it's the most widely adopted, but everything here applies regardless of which label you prefer. The point isn't the acronym. It's the outcome: buyers finding your business through AI.

Why businesses that move now will pull ahead

Right now, most businesses are still pouring their entire marketing budget into traditional SEO and Google Ads. The percentage of businesses actively optimizing for AI search is still in single digits. That gap won't last. By late 2026, the early movers will have established their entities in AI training data, built up third-party mention profiles, and locked in the authority signals that AI engines rely on to make recommendations.

The businesses that wait will find themselves in the same position as companies that ignored SEO in 2005. By the time they catch up, the competition has a multi-year head start in domain authority, content depth, and brand recognition in the AI's training data. AI engines tend to reinforce what they already "know." If ChatGPT already associates your competitor with your product category, unseating them gets harder with every passing month.

There's also a compounding effect at play. When ChatGPT recommends your business, that recommendation generates traffic, which generates reviews and mentions, which get picked up in future training data and web search results, which makes ChatGPT even more likely to recommend you next time. It's a flywheel, and the earlier you start spinning it, the faster it goes.

Chapter 2: How ChatGPT decides which brands to recommend

If you want ChatGPT to recommend your business, you need to understand how it picks winners. This isn't a black box, exactly. We can't see the full algorithm, but enough research has been done at this point to map out the major factors. Once you understand the machine, you can feed it what it's looking for.

ChatGPT's recommendation engine is a mix of baked-in knowledge from its training data and real-time information it pulls from the web when browsing mode is active. Both channels matter, and they work together in ways that most business owners don't yet appreciate.

Training data, entity recognition, and RAG

ChatGPT was trained on an enormous volume of text from the internet, books, articles, forums, and other sources. During that training process, it absorbed associations between words, concepts, brands, and categories. If your business was mentioned frequently in the training data, in positive contexts, and in connection with your product category, ChatGPT "knows" about you and is more likely to mention you when someone asks a related question.

Entity recognition is the mechanism behind this. AI engines don't just match keywords. They identify entities (people, companies, products, locations, concepts) and the relationships between them. When someone asks "What's a good CRM for real estate agents?", ChatGPT isn't searching for pages that contain those exact words. It's reasoning about the entity "CRM," the entity "real estate agents," and which CRM entities it associates with that audience based on everything it's absorbed.

Then there's Retrieval-Augmented Generation (RAG). When ChatGPT has web browsing enabled, it doesn't rely only on its training data. It searches the web in real time, retrieves relevant pages, scores them by how well they match the user's query (using a measure called cosine similarity), and incorporates that information into its answer. This is why your website content, your third-party mentions, and your structured data all matter: they're what ChatGPT retrieves and reads before answering.

The citation hierarchy

Research by Onely and BrightEdge has mapped out what types of content ChatGPT leans on most heavily when deciding which brands to recommend. The breakdown looks roughly like this:

The big takeaway: your own website accounts for only about 15% of the signal. The rest comes from what other people say about you. This is a fundamental difference from traditional SEO, where on-site optimization often accounted for the majority of your ranking power. In AI search, third-party validation is the primary currency.

Brand mentions vs. citations

BrightEdge's data shows that ChatGPT mentions brands 3.2 times more often than it cites them with a clickable link. This distinction matters more than most people realize. A "mention" is when ChatGPT says "You might want to look at Basecamp" in its answer. A "citation" is when it includes a hyperlinked source at the bottom of the response.

Mentions are the primary visibility mechanism. Most of the buyer influence happens in the body of the answer, where ChatGPT names specific businesses and explains why they're worth considering. Citations matter too (they drive direct traffic), but if you're only tracking whether ChatGPT links to your website, you're missing the bigger picture.

ChatGPT cites sources in 87% of its answers but only mentions brands in about 20.7%. That 20.7% figure tells you something important: the bar for getting mentioned is real. ChatGPT isn't dropping brand names casually. When it does mention a brand, it's because multiple signals have converged to make that brand the most relevant answer. Your goal throughout this book is to build those converging signals.

Content freshness and cosine similarity

One of the more surprising findings from recent research: 71% of the sources ChatGPT cites come from content published between 2023 and 2025. Older content, even if it ranks well on Google, gets significantly less weight in AI recommendations. ChatGPT (especially with browsing enabled) prioritizes recent information. This means a blog post you published last month can outweigh a competitor's cornerstone content from 2019.

Citation velocity matters too. Businesses that are mentioned frequently in content published in the past few weeks tend to get recommended more often than legacy brands that haven't had recent coverage. It's not enough to build authority once and coast. You need ongoing mentions.

At a technical level, when ChatGPT's browsing mode retrieves web pages, it scores them using cosine similarity, a mathematical measure of how closely a document's meaning matches the user's query. The higher your content scores on this measure, the more likely it gets retrieved and used in the answer. This is why writing content that directly mirrors the questions your customers ask (rather than generic industry content) is so effective for AEO.

Chapter 3: Content structure that ChatGPT can actually use

Most business websites are written for humans browsing, not for AI extracting. That worked fine when Google just needed to rank your page in a list. But ChatGPT needs to pull specific facts, comparisons, and recommendations out of your content and weave them into a conversational answer. If your content is buried in marketing fluff, vague claims, or walls of unstructured text, the AI will skip over you and pull from someone whose content is easier to parse.

This chapter covers how to structure your website content so that ChatGPT can actually find, understand, and use what you've written. None of this is complicated, but it requires rethinking how you organize information on your pages.

Writing for extraction, not just reading

When a human reads your website, they scan, skip around, and form impressions. When ChatGPT reads your website (via its RAG system), it's looking for extractable facts. It wants clear statements it can pull into an answer. "We provide world-class solutions for forward-thinking businesses" gives it nothing to work with. "We provide bookkeeping services for restaurants with 5-50 employees, starting at $299/month" gives it a specific fact it can use when someone asks about restaurant bookkeeping.

The first rule of AEO-friendly content: state things plainly. Every service page, product page, and key landing page should contain direct, factual statements about what you do, who you do it for, what it costs (if applicable), and what makes you different. Put the most important statement in the first two paragraphs of the page. ChatGPT (and other AI engines) weight the opening of a page more heavily than content buried at the bottom.

Think of it this way: if someone asked "What does [your business] do?", could ChatGPT assemble a clear, accurate answer from the first 200 words of your homepage? For most business websites, the answer is no. The first 200 words are usually a tagline, a stock photo, and some aspirational language. Fix that, and you've already jumped ahead of 90% of your competitors in AI readability.

The "direct answer" pattern

On every page that answers a question your customers commonly ask, start with the direct answer. Don't save the punchline for the end. Don't build up to your recommendation with three paragraphs of context. Open with the answer, then provide the supporting detail below it.

Here's what this looks like in practice. Suppose you run a pest control company and you have a page about termite treatment:

The FAQ-first content strategy

FAQ sections are the single highest-impact content format for AEO. The reason is straightforward: people ask AI engines questions, and FAQ content is literally structured as questions and answers. The match between format and use case is nearly perfect.

But we're not talking about the generic FAQ pages most businesses have, the ones with three questions like "What are your hours?" and "Do you offer free shipping?" Those are fine for customer service, but useless for AEO. What works is building substantial FAQ content around the questions your potential buyers actually ask before they purchase.

Go beyond the obvious. A good AEO FAQ section for an HVAC company wouldn't just cover "What brands do you carry?" It would answer questions like "How do I know if my AC unit needs to be replaced or just repaired?", "What's the real difference between a 14 SEER and a 20 SEER unit in terms of my electric bill?", and "Should I get a heat pump if I live in a climate where it drops below 20 degrees?" These are the questions buyers type into ChatGPT, and the business whose FAQ answers them clearly will be the one ChatGPT cites.

Where to place FAQ content

Don't limit your FAQs to a single /faq page. The highest-performing approach is to add relevant FAQ sections to every service page, product page, and core blog post on your site. Each page's FAQ should contain 5-10 questions directly related to that page's topic. A page about your residential plumbing services should have FAQs about residential plumbing. Your commercial plumbing page should have different FAQs about commercial scenarios.

This approach gives ChatGPT multiple entry points to discover and cite your content. When someone asks about residential plumbing costs, ChatGPT retrieves your residential page with its matching FAQs. When someone asks about commercial plumbing for office buildings, it retrieves that page instead. One FAQ page for your entire site can't do this.

Heading hierarchy and page architecture

AI engines use heading tags (H1, H2, H3) to understand the structure and hierarchy of your content. A well-structured page tells the AI exactly what each section is about before it reads the body text. A poorly structured page (or one that uses headings for visual styling rather than semantic meaning) forces the AI to guess, and it often guesses wrong.

The rules are simple. Each page gets one H1 that clearly states the page topic. H2s break the page into major sections. H3s break those sections into subsections. Every heading should be descriptive enough that someone reading only the headings would understand the page's full scope.

Notice the difference. The weak headings tell the AI almost nothing about what the page covers. The strong headings tell it the exact service, the location, and the specific subtopics. When someone asks ChatGPT "How much does commercial cleaning cost in Dallas?", the page with the strong heading structure is vastly more likely to get retrieved and cited.

Authorship, dates, and credential signals

ChatGPT gives more weight to content that has clear authorship, recent publication dates, and visible credentials. This is partly because its training data (and its RAG retrieval) have been shaped by quality signals that major platforms like Google have long used, and partly because well-attributed content is simply more trustworthy from a factual standpoint.

Every blog post and content page on your site should include the author's name and a brief bio that explains their expertise. Include their credentials if they have relevant ones. Add a publication date and an "updated" date when you revise content. Create dedicated author bio pages that link to the author's LinkedIn, relevant publications, or other authority indicators.

For service and product pages, make sure your About page clearly states who runs the company, their background, and any relevant certifications or accreditations. AI engines use this information for entity recognition, the process of connecting your brand to a real person with verifiable expertise.

Chapter 4: Schema markup and structured data for AI engines

Structured data is how you tell AI engines what's on your page in a language they can read without guessing. When you add schema markup to your website, you're labeling your content with machine-readable tags: this is a product, this is its price, this is a review, this person is the author, this organization is the publisher. AI engines can then pull these facts with confidence instead of trying to interpret your marketing copy.

The data backs this up. Content with proper schema markup is 2.5 times more likely to appear in AI-generated answers compared to equivalent content without it. Yet most small and medium businesses still don't have any schema markup beyond the basics their website platform generates automatically. This chapter shows you what to add and how.

JSON-LD: the non-negotiable standard

There are three formats for adding structured data to a web page: Microdata, RDFa, and JSON-LD. Forget the first two. JSON-LD (JavaScript Object Notation for Linked Data) is the format that Google officially recommends, that Bing requires for most rich result types, and that AI engines like ChatGPT process most reliably. It's also the easiest to implement because it goes in a single script tag in your page's head section, separate from your HTML content.

A basic JSON-LD block for a local business looks like this:

That block takes five minutes to add and immediately tells every AI engine your business name, location, contact info, price range, hours, and social profiles. Without it, the AI has to parse your entire page and try to figure all of that out from context.

FAQPage schema and why it outperforms everything else

FAQPage schema has emerged as the highest-performing structured data type for AI search optimization. The reason connects directly to Chapter 3: AI engines respond to questions, and FAQPage schema wraps your questions and answers in a format that's trivially easy for them to parse and cite.

When you add FAQPage schema to a page that already has FAQ content, you're essentially double-signaling. The AI can see the FAQ in your visible page content and also in the structured data, which makes it far more confident about the accuracy and relevance of those answers.

Every service page, product page, and major content page on your site should have FAQPage schema. It's one of the few AEO tactics with zero downside and consistently measurable upside.

Entity depth: connecting your brand to the knowledge graph

Here's where schema markup gets genuinely powerful for AEO. AI engines don't just read individual schema blocks in isolation. They follow the connections between entities. When your Product schema links to your Organization schema, which links to your Founder schema (a Person), the AI builds a connected picture of your brand's identity.

This is called entity depth, and it's one of the strongest signals you can build. The chain looks like this:

Each level in this chain adds confidence for the AI. A product listed on a random page is just a name. A product linked to a manufacturer, which is linked to an organization with a real address and a founder with a LinkedIn profile and media mentions, that's a verified entity the AI can trust and recommend.

Google's March 2026 schema update

In March 2026, Google rolled out an update that tightened its rules around schema markup. The update cracked down on pages where the schema described content that wasn't the primary focus of the page. For example, a general blog post with Product schema for items that were only briefly mentioned (rather than being the page's main subject) could get its rich results stripped.

The update also increased the weight of schema as an entity verification signal. In plain terms, Google (and by extension, the AI engines that reference Google's knowledge graph) now puts more trust in schema that accurately represents a page's content and more penalty on schema that overstates or misrepresents it.

What this means for you: be honest with your schema. Only mark up content that genuinely exists on the page. Don't add FAQPage schema to a page that doesn't have visible FAQ content. Don't add Product schema with a price that doesn't appear on the page. The upside of accurate schema has increased, and the downside of misleading schema has too.

Chapter 5: The llms.txt file and technical AI-readiness

There's a new file you should add to your website's root directory, alongside your robots.txt and sitemap.xml. It's called llms.txt, and it's essentially a table of contents for AI. While the specification is still evolving, early adopters are already seeing measurable benefits, and adding one takes about 30 minutes.

But llms.txt is just one piece of the technical puzzle. This chapter covers the full set of technical optimizations that make your site easy for AI engines to crawl, understand, and cite. Most of these aren't glamorous, but they're the foundation that everything else in this book builds on.

What llms.txt is and how to create one

The llms.txt file is a Markdown-formatted document that sits at yourdomain.com/llms.txt. It provides AI engines with a structured overview of your site: what your business is, what your key pages are, and how the content is organized. Think of it as a guided tour of your website, written specifically for an AI audience.

The specification (maintained at llmstxt.org) calls for a specific structure: an H1 heading with your project or business name, a blockquote with a short summary, optional additional context paragraphs, and then organized link sections using H2 headings with bullet-pointed links and descriptions.

Crawlability and clean HTML

AI engines access your content through web crawling, just like Google does. If your site is slow to load, blocks crawlers, or wraps its content in JavaScript that requires rendering, AI engines will either skip you or misread your content.

The basics matter here: your pages should load in under 3 seconds, your content should be in the HTML source (not loaded dynamically via JavaScript after page load), and your navigation should use standard HTML links that a crawler can follow. If your site is built on a JavaScript framework like React or Vue, make sure you're using server-side rendering (SSR) or static site generation (SSG) so crawlers see the full content.

Clean HTML also means proper semantic markup. Use paragraph tags for paragraphs, list tags for lists, table tags for tabular data. Don't build your page layout entirely out of div tags with CSS. AI engines use HTML semantics to understand the type and structure of content on the page.

Robots.txt, sitemaps, and AI crawler access

Your robots.txt file controls which crawlers can access which parts of your site. There's been debate in the industry about whether to block AI crawlers (like GPTBot, which is OpenAI's crawler) or let them in. For AEO purposes, the answer is clear: let them in.

If you block GPTBot in your robots.txt, you're telling ChatGPT's crawler that it's not allowed to read your content. That means ChatGPT's browsing mode can't access your pages when generating answers, and your content won't be included in future training data. You're essentially opting out of AI recommendations.

Check your robots.txt file right now (yourdomain.com/robots.txt). If you see any of these user agents being blocked with "Disallow: /", remove those blocks. If you don't have a robots.txt file, that's fine. By default, all crawlers are allowed.

For your XML sitemap, make sure it lists all your important pages and is submitted to Google Search Console and Bing Webmaster Tools. While AI crawlers don't all use sitemaps directly, the search engines that feed data to AI engines do.

Building an AI-friendly site architecture

Site architecture is how your pages are organized and linked together. For AI readability, the best structure is a clear hierarchy: homepage links to category pages, category pages link to specific service or product pages, and every page is reachable within three clicks from the homepage.

Each major topic or service should have a "hub" page that links to more detailed sub-pages. This creates topical clusters that AI engines recognize. When ChatGPT sees that your site has a main "Commercial Cleaning" hub page that links to sub-pages about office cleaning, medical facility cleaning, retail cleaning, and warehouse cleaning, it understands that your business has depth in commercial cleaning. That depth makes it more confident in recommending you for commercial cleaning queries.

Chapter 6: Building authority that AI trusts

Remember the stat from Chapter 2: roughly 85% of the signals AI uses to recommend brands come from third-party sources, not from your own website. That makes this chapter arguably the most important one in the book. You can have the most beautifully structured website with perfect schema markup and an llms.txt file, but if nobody else is talking about you, AI engines won't have enough confidence to recommend you.

Building authority for AI search is different from building it for traditional SEO. It's not just about backlinks (though those still help). It's about being mentioned by name, in positive contexts, across sources that AI engines recognize as trustworthy. Let's break down how to do that.

Digital PR for AI visibility

Digital PR, the practice of getting your business mentioned in online publications, news outlets, and industry media, has become the single most effective way to build AI authority. Research from GenOptima's monitoring data shows that 61% of the signals informing AI's understanding of brand reputation come from editorial media sources.

The mechanism is straightforward. When a respected industry publication writes "For small restaurant owners looking to simplify their books, FreshBooks stands out for its tip-tracking feature," that sentence gets indexed. AI engines encounter it during training and during web search retrieval. It builds the association between FreshBooks, restaurant accounting, and a positive recommendation. Multiply that across several publications, and the association becomes strong enough for ChatGPT to confidently recommend FreshBooks when asked about restaurant accounting software.

What kinds of PR coverage work best for AEO? The most effective types are "best of" lists and roundup articles in your industry ("The 10 best [product category] for [audience]"), expert commentary and quotes in news articles about your industry, case studies published in industry publications, and thought leadership articles where you're positioned as a subject matter authority.

Getting PR coverage without a huge budget

You don't need a $10,000/month PR agency to get mentioned in relevant publications. Here's what actually works for small and medium businesses:

First, become a source. Sign up for journalist query platforms like HARO (Help a Reporter Out), Qwoted, and SourceBottle. Journalists post requests for expert quotes and sources daily. When you see a query related to your industry, respond with a thoughtful, specific answer that includes your name, title, and business name. Even one mention in a mid-tier publication can significantly boost your AI visibility.

Second, pitch guest articles to industry blogs and trade publications. Most industry publications accept contributed articles from practitioners. Write about something you know well, include real numbers and examples from your experience, and make sure your author bio clearly states your business name and what you do.

Third, seek out "best of" list inclusion. Many of these lists are actively maintained and accept submissions. Find the lists that rank for your product category on Google, then look for submission guidelines, nominate your business for consideration, or reach out to the author directly.

Wikipedia, Wikidata, and entity optimization

Wikipedia and Wikidata are two of the most powerful authority signals for AI engines. ChatGPT's training data includes Wikipedia extensively, and the structured data in Wikidata feeds directly into knowledge graphs used by Google, Bing, and other platforms that AI engines reference.

If your business or its founder is notable enough for a Wikipedia article (and "notable" in Wikipedia terms means having significant coverage in independent, reliable sources), getting a Wikipedia presence can dramatically increase your AI visibility. But Wikipedia has strict rules about neutrality, sourcing, and conflicts of interest. You cannot write your own Wikipedia article, and you should not hire someone to do it under the guise of being an independent editor. What you can do is build up the independent media coverage (through digital PR) that would make a Wikipedia article justifiable, and then let the Wikipedia community notice.

Wikidata is more accessible. It's Wikipedia's structured data counterpart, and you can create entries for your business, your products, and key people without the same notability requirements. A Wikidata entry gives your brand a stable identifier that knowledge graphs can reference.

Google Knowledge Panel and knowledge graph presence

Google's Knowledge Panel is that information box that appears on the right side of search results when you search for a well-known entity. Having one means Google has recognized your business as a distinct entity, and AI engines that reference Google's knowledge graph inherit that recognition.

To trigger a Knowledge Panel, you need consistent business information across the web (name, address, website, social profiles), a claimed and verified Google Business Profile, schema markup on your website (Chapter 4), and mentions in authoritative sources (digital PR). It's not something you apply for. Google generates it automatically when it has enough confident data about your entity.

Once you have one, claim it through Google's knowledge panel verification process. This lets you suggest edits and ensure the information is accurate. An accurate Knowledge Panel feeds accurate information to AI engines, which feeds accurate recommendations to users.

Industry directories and review platforms

Directory listings are the easiest authority signals to build. They take minutes to create, they're usually free, and AI engines actively reference them. The key is choosing the right directories for your industry and keeping your information consistent across all of them.

The critical detail: your business name, address, phone number, and website URL must be identical across every directory. Even small inconsistencies (like "St." on one listing and "Street" on another) can confuse entity recognition. Pick one format and use it everywhere.

Chapter 7: Review and reputation management for AI citations

Online reviews account for about 16% of what influences ChatGPT's brand recommendations, which puts them third in the citation hierarchy behind authoritative lists and awards. But reviews punch above that 16% weight because they also affect the other signals. A business with strong reviews is more likely to appear on "best of" lists, more likely to win awards, and more likely to be referenced positively in media coverage. Reviews are the amplifier that makes everything else work harder.

This chapter covers how AI engines process reviews, which platforms matter most, and how to systematically build a review profile that makes ChatGPT confident in recommending you.

How AI engines weigh online reviews

AI engines don't just look at your star rating. They read the text of reviews and extract patterns. They pick up on what specific features or qualities reviewers mention, whether the sentiment is positive or negative, how recently the reviews were posted, and whether the same themes appear across multiple platforms.

When someone asks ChatGPT "What's the best accounting software for freelancers?", and ChatGPT recommends FreshBooks, the reasoning it provides often mirrors the language from actual reviews: "easy to use," "good invoicing," "responsive support." It's pulling these characterizations from review text, not just from the company's marketing. This means the specific words your customers use in their reviews directly affect how ChatGPT describes and recommends your business.

Volume matters too, but with diminishing returns. Having 10 reviews vs. 0 is a massive difference. Having 500 vs. 200 matters less. What doesn't diminish is recency. A business with 200 reviews, all from two years ago, looks dormant to an AI engine. A business with 50 reviews, 15 of them from the past three months, looks active and current. Aim for a steady flow of new reviews rather than a one-time batch.

Platforms that matter most

Not all review platforms carry equal weight with AI engines. The platforms that matter most depend on your business type, but some general principles apply:

The best approach is to focus your review-building efforts on 2-3 platforms that are most relevant to your business type. Don't spread yourself across seven platforms with a few reviews on each. Concentrate your social proof where it counts most for your category.

Systematic review generation

Getting reviews consistently requires a system, not just occasional asks. Here's what works without being pushy or violating platform guidelines:

Build a review request into your delivery process. After you complete a service, deliver a product, or reach a milestone with a client, send a short email (or text) with a direct link to your preferred review platform. The timing matters. Ask within 24-48 hours of a positive interaction, when the experience is still fresh. Don't ask during a complaint or while an issue is unresolved.

Make it specific. "Could you leave us a review?" gets low response rates. "Would you mind sharing what your experience was like with our kitchen remodel? Here's a direct link to leave a Google review" gets much higher response rates because it tells the customer exactly what to write about.

For B2B companies, the most effective approach is a personal ask from the person who manages the relationship. An email from your account manager saying "We just wrapped up your Q1 reporting project and I'm glad the numbers looked strong. Would you be open to sharing a quick review on G2? It takes about 3 minutes" converts far better than an automated survey email.

Awards, certifications, and accreditation signals

Awards and accreditations account for about 18% of ChatGPT's citation signals, which is actually more than reviews. They work because they're third-party validation that a trusted organization has vetted your business and found it worth recognizing.

The types that carry the most weight for AI: industry-specific awards (like a "Best New Product" award from a trade association), platform awards from review sites (G2 Leader Badge, Capterra Shortlist), professional certifications relevant to your field, and BBB accreditation.

Many of these are achievable for small businesses. G2 badges, for instance, are calculated based on review volume and ratings. If you build enough strong reviews on G2, you'll qualify for badge categories automatically. Trade associations often have award programs that don't require you to be a giant corporation. Look for what's available in your specific industry.

Chapter 8: Content marketing and thought leadership for AI discovery

Content marketing for AEO follows a different logic than content marketing for Google SEO. The Google playbook was: publish frequently, target long-tail keywords, build topical authority through volume. The AEO playbook is: publish less often but with more substance, focus on depth over breadth, and make sure your content gets mentioned and referenced by others. AI engines prefer brands that demonstrate focus and depth, and they can tell the difference between a hundred thin blog posts and ten authoritative pieces that actually teach something.

This chapter covers how to adjust your content strategy for AI discovery, including channels that most businesses overlook.

Fewer, deeper pieces vs. high-volume content

There's a counterintuitive finding in the AEO data: publishing fewer but stronger pieces around one subject builds AI confidence in your brand faster than covering many topics briefly. The reason goes back to how entity recognition works. When AI engines see that your business has published five in-depth articles about restaurant inventory management (each with specific data, examples, and actionable detail), they associate your brand strongly with that topic. When they see your business has published 50 articles about 50 different small business topics, each one a surface-level overview, the association is weak across all of them.

This doesn't mean you should only write about one thing. It means you should pick 3-5 core topics that directly relate to your business, and go deeper than anyone else on each of them. A 3,000-word guide that answers every question a buyer could have about your specific niche is worth more for AEO than thirty 500-word blog posts about tangentially related topics.

The "definitive resource" approach

For each of your core topics, create one piece of content that aims to be the most comprehensive resource available. This means including original data or examples from your own business (AI engines notice unique information that doesn't appear on other sites), specific numbers, timelines, and case studies, practical tools like checklists, templates, or calculators, and answers to the 15-20 questions buyers commonly ask about this topic. When this content is genuinely the best resource on its topic, other sites link to it and mention it, which creates the third-party signals that drive AI recommendations.

YouTube and video as an AI visibility factor

Here's one that catches people off guard: YouTube mentions and branded video content are among the top factors correlated with AI brand visibility. Research from multiple AEO studies has found that branded web mentions and YouTube mentions are the top two factors correlating with AI visibility in ChatGPT, Google AI Overviews, and similar platforms.

The connection makes sense when you think about it. YouTube is the second-largest search engine, and its transcripts are part of the training data for most major AI models. When someone publishes a video titled "Best CRM for Real Estate Agents" and mentions your product by name, that's a signal AI engines pick up. When multiple videos do this, the signal gets strong.

You have two angles here. First, create your own YouTube channel with content relevant to your product category. Even simple videos (screen recordings of your product, how-to tutorials, customer interviews) build your entity presence on the platform. Second, get mentioned in other people's videos. This is the YouTube equivalent of digital PR, and it often comes naturally if you're doing the other authority-building work in this book (reviews, awards, industry engagement).

LinkedIn, podcasts, and guest content

LinkedIn has become surprisingly important for AEO. AI engines process LinkedIn content during training, and LinkedIn's domain authority means content published there (articles, posts, company page descriptions) gets picked up in web search retrieval too. A strong LinkedIn presence for both your company and its key people builds entity recognition and reinforces the associations AI engines make between your brand and your expertise areas.

For LinkedIn, the approach that works best for AEO is regular posting about your specific area of expertise (not generic business advice), publishing long-form articles that demonstrate depth, engaging with others in your industry (which creates additional entity connections), and keeping your company page description clear, factual, and keyword-aligned with how customers describe your product category.

Podcast appearances work similarly. When you appear as a guest on podcasts in your industry, the show notes and transcripts create additional web content that mentions your brand, your expertise, and your business in context. Podcast content is also increasingly included in AI training data. Being a guest on 5-10 relevant podcasts can meaningfully boost your brand mentions across the web.

Citation velocity: why recency beats volume

We touched on this in Chapter 2, but it's worth expanding here because it has direct implications for your content calendar. Citation velocity is how frequently your brand gets mentioned across the web within a recent time window. Businesses that are mentioned frequently in content from the past few weeks tend to outperform legacy brands with older, static authority.

This means your content marketing and PR efforts need to be ongoing, not one-time campaigns. A steady rhythm of new content, new PR mentions, new podcast appearances, and new reviews signals to AI engines that your brand is current, active, and relevant. A brand that was heavily covered in 2023 but has had no new mentions since will gradually fade from AI recommendations as fresher content about competitors takes its place.

Chapter 9: Monitoring, measuring, and improving your AI visibility

You can't optimize what you can't measure. Traditional SEO had mature analytics tools from the start: Google Analytics, Search Console, rank trackers. AI visibility tracking is newer, but a growing ecosystem of tools now exists to tell you exactly how your brand shows up in AI-generated answers, how you compare to competitors, and where your gaps are. This chapter walks you through the tools, the metrics, and the process for turning data into action.

The AI visibility toolbox

Several specialized platforms have launched in 2025-2026 specifically to track brand visibility in AI search. Here's what's available and what each one does well:

If you're just starting out and budget is tight, HubSpot's free AEO Grader is a reasonable first step. It won't give you ongoing monitoring, but it'll score your current readiness and point you toward the biggest gaps. For ongoing tracking, Otterly AI and AIclicks offer the broadest coverage of AI platforms at reasonable price points.

Setting up tracking from scratch

Even without paid tools, you can set up basic AI visibility tracking manually. It takes about 30 minutes per week and gives you enough data to guide your optimization efforts.

Start by building a list of 20-30 prompts that represent the questions your potential customers would ask AI engines. These should cover your product category, your specific location (if local), comparison queries (e.g., "X vs. Y"), and recommendation queries ("best X for Y"). Run each prompt through ChatGPT once per week, and record: whether your brand was mentioned, where it appeared in the answer (first recommendation, listed among several, or absent), what ChatGPT said about your brand, and what competitors were mentioned.

Track these in a simple spreadsheet. Over weeks and months, you'll see patterns: which prompts you're winning, which you're losing, and how your visibility changes as you implement the strategies in this book.

Competitor benchmarking in AI search

AI visibility is relative. You're not trying to hit an absolute score. You're trying to appear more often, and more favorably, than your competitors for the prompts that matter to your business. That's why competitor benchmarking should be part of your regular tracking.

For each of your tracking prompts, note which competitors ChatGPT mentions and what it says about them. Look for patterns. Maybe a competitor consistently gets recommended for a specific use case that you also serve. That tells you to create more content, reviews, and PR coverage around that use case. Maybe your competitor is never mentioned for a certain topic where you have genuine expertise. That's an opportunity to claim territory they haven't.

The most useful competitive metric is share of voice: across your 30 tracking prompts, what percentage of the time does ChatGPT mention you vs. each competitor? Track this weekly and you'll see the direct impact of your AEO efforts over time.

The audit-optimize-measure loop

AEO is not a set-it-and-forget-it task. It follows a continuous improvement loop:

Audit means running your tracking prompts, checking your schema, reviewing your directory listings, and noting what's changed since last month. Optimize means taking action on the gaps: publishing new content, sending PR pitches, requesting reviews, fixing technical issues. Measure means tracking the results of those actions in the next audit cycle.

Monthly is the right cadence for most businesses. AI engines update their training data and search indexes on varying schedules, so changes you make today might not show up in recommendations for 2-6 weeks. Weekly prompt tracking catches fast changes, but the full audit-optimize-measure cycle works best as a monthly discipline.

Chapter 10: Your 90-day AEO action plan

Everything in this book comes down to execution. You can understand every concept perfectly and still get zero results if you don't actually do the work. This chapter gives you a concrete, week-by-week plan for your first 90 days of AEO implementation. It's designed to be achievable for a small team (or even a solo business owner) investing 5-10 hours per week in AEO alongside their regular work.

The plan is structured in three phases: foundation (days 1-30), authority building (days 31-60), and scale and optimization (days 61-90). Each phase builds on the one before it, so don't skip ahead.

Days 1-30: Foundation and quick wins

The first month is about getting the technical foundation right and picking up the easy wins that don't require outside help or long lead times.

Days 31-60: Authority building

With the foundation in place, month two shifts to building the third-party signals that drive most AI recommendations.

Days 61-90: Scale and optimize

Month three is about doubling down on what's working and building the ongoing systems that will sustain your AI visibility long-term.

Future-proofing beyond ChatGPT

ChatGPT is the biggest player in AI search right now, but it's not the only one. Perplexity AI has grown rapidly as a dedicated AI search engine. Google's AI Overviews are reshaping how Google results work. Bing Copilot, powered by the same technology as ChatGPT, is built into Microsoft's browser and operating system. Meta's AI is integrated into Facebook, Instagram, and WhatsApp.

The good news is that almost everything in this book applies across all of these platforms. The principles are the same: clear, structured content; strong entity recognition; third-party authority signals; fresh mentions; and consistent presence across the web. A brand that's optimized for ChatGPT recommendations is also well-positioned for Perplexity, Google AI Overviews, and whatever comes next.

The things to watch going forward: how AI engines handle advertising (ChatGPT has begun testing sponsored recommendations, which could change the landscape for paid visibility), how training data and crawling policies evolve (some publishers are restricting AI access, which could create gaps in coverage), and how user behavior continues to shift (the more people trust AI answers, the more important AEO becomes relative to traditional SEO).

The businesses that will thrive are the ones that treat AEO as an ongoing discipline, not a one-time project. Keep the flywheel spinning: create content, build authority, earn mentions, monitor results, optimize, and repeat. The compounding effect of consistent AEO work will widen the gap between you and competitors who are still focused entirely on Google rankings while the world moves on.