She’s on her lunch break, phone in hand, and she already knows what she wants, kind of. Something spicy but not raunchy. Enemies-to-lovers, maybe. Fast-paced enough for the train ride home, but with enough emotional depth to keep her up at night. She doesn’t open Google. She doesn’t scroll Goodreads. She opens ChatGPT and types: “Recommend a spicy romance novel, enemies-to-lovers, fast-paced, good for travel reading, similar to what’s trending on BookTok.”
The AI responds in seconds. It names three titles with a sentence of context about each. One of those books gets bought before she finishes her lunch. The others don’t exist in her world.
That’s not a hypothetical. It’s happening thousands of times a day across romance, literary fiction, nonfiction, cozy mystery, and every other genre your catalog touches. How readers discover books has quietly undergone one of its biggest shifts in decades, and most independent publishers haven’t caught up yet.
That shift has a name: generative engine optimization, or GEO. And understanding it may be the most important thing you do for your titles this year.
What GEO Is (and Isn’t)
You already know SEO, search engine optimization, the craft of making your website and content show up in Google results. GEO is the next layer built on top of it.
GEO is the emerging practice of optimizing your content so that AI tools, including ChatGPT, Perplexity, and Google’s AI Overviews, choose to reference, cite, and recommend your books when readers ask for them.
The distinction matters because the mechanics are different. Traditional search engines return a list of links and let users decide where to click. Generative engines synthesize information from multiple sources into a single, conversational answer. Your books don’t just need to rank; they need to be described clearly and authoritatively enough for AI to present them confidently to a reader.
GEO doesn’t replace SEO. Think of it as a second floor built on the same foundation. Quality content, consistent information, and credible third-party mentions all feed both. But GEO adds specific requirements that traditional SEO never demanded.
Why the BookTok Reader Is Your Best Case Study
The romance and romantasy space is where GEO dynamics are most visible, because BookTok readers are already fluent in the language of AI-style queries. They don’t search for a title. They describe an experience. Spicy vs. cozy. Slow burn vs. instant connection. Beach read vs. deep immersion. These conversational prompts map almost perfectly onto how readers now talk to AI tools.
Here’s the problem for publishers: If the AI can’t find clear, specific, descriptive language about your title across the web, it won’t recommend it, even if the book is a perfect match. AI tools default to the most mainstream, widely discussed titles. Your emerging or backlist titles get passed over, not because they’re worse, but because they’re less legible to the AI. That’s a GEO problem, and it’s fixable.
What AI Actually Looks For
Most AI search tools use a process called retrieval-augmented generation. They pull relevant snippets from across the web in real time and synthesize them into a response. Think of it like an open-book exam. The AI has a massive library but only pulls the pages that directly answer the question.
Four things determine whether your content gets pulled:
Specific, descriptive language. Vague marketing copy is invisible to AI. “A gripping page-turner” tells an AI nothing. But “a psychological thriller with an unreliable narrator, perfect for readers who loved Freida McFadden” is citable. Every book description should read like a direct answer to a question a reader might actually ask.
Consistency across platforms. AI tools build “entity profiles” by reading across many sources. If an author’s name is spelled differently on Amazon than on Goodreads, or your book’s genre description differs between your site and its press kit, the AI moves on to something clearer. Consistency is a GEO signal.
Authority signals across the web. AI systems weigh credibility based on how many independent sources discuss your titles in context. Reader reviews, book blogger posts, author interviews, press coverage—all of it feeds the AI’s sense of whether your book is authoritative enough to recommend.
Technical accessibility. If AI crawlers can’t read your site, none of the above matters. Two quick wins: ensure your site isn’t blocking AI crawlers via robots.txt, and implement schema markup on book and author pages. Schema acts as structured metadata that tells AI systems what your content is, who created it, and how it should be categorized.
What Indie Publishers Can Do Right Now
The good news is that most GEO best practices are things you should already be doing. Here’s where to focus first.
1. Build and maintain author websites. The author website is one of the biggest winners of the AI era. It’s where you have full control over how your authors and titles are described, and it’s a primary source AI tools can crawl and cite.
2. Write for the questions readers actually ask. Every book page should answer the questions a reader might ask an AI: What tropes are in this book? Who is this for? What mood does it suit? What comparable titles does it sit alongside? Answer those questions plainly and near the top of the page.
3. Invest in genre-specific descriptors. Every category has its own vocabulary, and AI has learned it. “Slow burn,” “locked-room mystery,” “narrative-driven nonfiction.” These are the terms readers use in prompts. Use them consistently so your titles are clearly understood and surfaced.
4. Earn mentions in the right communities. Genre forums, BookTok-adjacent blogs, Goodreads, and newsletters shape what AI trusts and recommends. Advanced reader copies, community engagement, and targeted media placements now drive both reader buzz and AI visibility.
The Shift Is Already Happening
The conversation about GEO in publishing is just getting started, but the reader behavior driving it is already here. The BookTok user asking ChatGPT for her next read isn’t an edge case; she’s a signal of where discovery is heading.
Indie publishers have always competed by being nimble where larger houses are slow. This is one of those moments.
The path forward is incremental: Invest in a single author site, refine one high-quality book page, and secure one meaningful community mention. Repeat.
The publishers building this discipline now are the ones an AI will recommend a year from today.
Katia Herrera is the senior director of e-commerce and performance marketing at Sourcebooks. Since joining in 2017, she has executed performance marketing and online marketplace strategy across its imprints, spanning sponsored ads, DSP, and emerging media, with a focus on driving scalable discoverability and revenue across domestic and international markets. She currently leads the company’s GEO task force, working at the forefront of how AI is transforming book discovery globally.