SEO Is So Last Year: What Authors Need to Know About GEO

Just when I thought I had a handle on SEO, the rules changed.

For the past few years, author discoverability has been largely about search engine optimization: the right keywords, meta tags, page titles, and all the invisible machinery that helps Google decide whether your website deserves to show up on page one or page forty. I won't pretend I ever fully mastered it, but I at least understood the game. You identified the words your readers were searching for, you worked them into your site strategically, and you hoped the algorithm rewarded you for it.

Now there's a new game. It's called GEO, generative engine optimization, and it matters because the way people find things is quietly but fundamentally shifting. More and more, someone looking for their next read isn't typing into a search bar and scrolling through results. They're asking ChatGPT, or Gemini, or Claude: "I want a love story with some depth to it, not the usual formula. What should I read?" And an AI answers them directly, no list of links required, no page two to scroll to. Just an answer.

The question every author needs to be asking is whether their book is in that answer.

I've been thinking about this for a while, and over the past several weeks I've been doing something about it. The core insight is this: to have any chance of being recommended by an AI system, that system needs accurate, structured information about your work. Not just your title and genre, but the emotional experience readers have, the comparable authors, the themes, what your book is not. That last part turns out to be just as important as what it is. I spent real time building out a dedicated page on my website written specifically for AI systems to read and use. Plain language. Structured facts. Explicit corrections to the assumptions these models might otherwise make on their own.

One example: I included a clear note that my books involve parallel timelines, not time travel. It sounds like a small distinction, but for a recommendation engine trying to match a reader who specifically does not want time travel, it matters enormously. The AI needs to know the difference, and it can only know if you tell it.

It is a genuinely strange thing to write a page for an audience that isn't human. But given that this is increasingly how readers are discovering new authors, it felt worth the strangeness.

The first thing I did, before building anything, was ask the AI systems to tell me what they already knew. A knowledge check, essentially. What I found was instructive. The basic facts about my books were largely accurate, which was reassuring. What was missing was the emotional layer, the why-read-this that goes beyond plot summary and genre tags. Gemini reached for my professional credentials, the geochemist and hydrogeologist and marketing strategist, while ChatGPT delivered my author bio. Neither was wrong exactly, but they were telling different stories about the same person. It was a useful reminder that different systems prioritize different kinds of information, and that no single page or profile is going to reach all of them equally. I'm still working on injecting the emotional resonance into how these systems understand my books. That part is harder than getting the facts right.

One thing that did make a measurable difference was simply being consistent about how I identify myself. Using M. Jacqueline Murray rather than variations of my name has reduced the confusion with other people who share parts of it. My professional identity and my author identity haven't fully fused yet in the way AI systems present me, but I expect that will come with time and consistency.

Beyond the AI information page itself, I found a few other tactics worth pursuing. Structured metadata matters, the information attached to your books in distribution systems, library catalogues, and retail listings. Open Library, which feeds into WorldCat and library discovery systems, turned out to be worth updating with accurate, detailed records. Adding FAQ sections to my web pages, using code injection to make the structure machine-readable, is another layer that helps AI systems parse and prioritize information correctly. None of it is glamorous work. Most of it is invisible to human readers. But it builds the foundation that determines whether an AI, when asked for a recommendation, has enough to go on.

I would never claim to be the most technologically current person in any room I've ever been in. But I've always believed that understanding new technology, even imperfectly, is better than ignoring it until you're hopelessly behind. I watched the internet reshape not just publishing, but research, access to information, and the entire way human beings communicate with each other. I used email before most people had heard the word, back when the University of Waterloo's computer occupied an enormous cavern in the center of the MC Building. I lived through the shift from landlines to mobile phones, from card catalogues to Google Scholar, from physical maps to GPS. I've seen enough technological tipping points to recognize one when it's happening.

AI is one of those tipping points. And it's moving faster than most of the others.

I'm not suggesting every author needs to become an AI expert. I'm certainly not one. But I am suggesting it's worth paying attention, worth spending a few hours understanding how these systems work and what they need from you in order to work in your favour. The authors who figure that out now will have a head start on the ones who look up in a couple of years and wonder why their discoverability has quietly eroded.

For what it's worth, the process of building my AI information page also turned out to be useful in a completely unexpected way. Writing structured, precise descriptions of my books for a non-human audience forced me to articulate things I'd been saying vaguely for years. What is the emotional experience of reading my books? What do they share with Jodi Picoult's work, and where do they diverge? What would make someone who loved one of them immediately reach for the next? Those are good questions for any author to sit with, regardless of who or what is eventually going to read the answers.

The game keeps changing. I intend to keep playing.