A reader who is looking for their next thriller no longer always types "best new thriller authors 2026" into Google. A meaningful slice of them ask ChatGPT, scan the AI Overview that appears at the top of a Google search, use Perplexity, or post the question to a community thread that an AI then scrapes and summarises. The behaviour is real, and it has implications for author SEO that most of the advice currently circulating has not caught up with.
I want to start by saying what I think is true. Author SEO basics are not obsolete. Clean metadata, a well-structured author website, a focused keyword on each book page, descriptive image alt text, and useful internal linking still produce a measurable lift in discoverability. I’ve watched author sites double their organic traffic in 30 days from those fundamentals alone. None of that is going anywhere. The layer above it has shifted, and authors who treat SEO purely as a search-engine problem are leaving a real visibility gap on the table.
Here is what I want to walk through. What AI search is actually doing differently. What that changes about how an author website should function. What stays the same. And where I think the work should sit on your priority list, especially if you only have a few hours a month for it.
How AI search differs from keyword search
Traditional SEO is a popularity contest with a vocabulary requirement. You target a phrase, you optimise around that phrase, Google ranks the page, the reader clicks and lands on your site. The reader does the work of evaluating the result.
AI search compresses that step. A reader asks "who writes contemporary fantasy with romance subplots and a slow-burn arc, set in a real-world city" and the AI produces a list of three to seven names, often with a sentence of context for each. The reader does not click ten links. They click one, maybe two, if any.
What gets included in that list is decided by a different mechanism than search ranking. The model is pulling from training data, recent indexed content, structured information it can parse, and increasingly, real-time retrieval from web sources. It is not ranking based on backlinks and on-page keywords in the same way. It is making judgments about which authors fit the described shape.
For an author, that means two things. First, the question of whether you appear at all in those answers depends on whether the AI has been given enough specific, structured information about your work to place you correctly. Second, you have less direct control over what gets said about you than you do over your own search rankings. The shift is from "where do I rank" to "what does the model know about me, and can it describe me accurately."
What this changes about your website
Your author website was already supposed to function as the single source of truth on you and your books. AI search makes that more important, because the website is increasingly the source the AI is summarising from. The implications are practical.
Your About page should read like a source document, not a personal essay. That does not mean stripping the personality out of it. It means including the kind of factual, declarative content the model can lift: what genre you write, what subgenres specifically, what tropes recur in your work, what readers have compared your books to, where your books are sold, where your reader community lives. Most author About pages I audit are written for a reader who already knows they like you. AI search rewards About pages written for someone who does not know who you are yet.
Each book page should describe the book in terms that match how readers ask for it. That sounds obvious, but most book pages I audit are written as marketing copy aimed at conversion. They use language like "you’ll be swept away by" and "a story that will stay with you long after the last page." That language is fine for a reader who has already arrived on the page. It is invisible to an AI looking for tropes, genre markers, comparable authors, and structural details. A book page that says "enemies-to-lovers, slow-burn, dual-POV romantasy with a morally grey heroine, set in a city under siege" gets parsed correctly. A book page that says "a tale of love against the odds" does not.
Schema markup matters more than it did. Book schema, author schema, and review schema all give AI models structured data they can use. Most indie author websites have none of this. Adding it is a one-time investment.
What does not actually change
I want to name this clearly because there is a strong temptation right now to treat every long-standing recommendation as obsolete. It is not.
Newsletter lists are still the most durable visibility asset an author can build, which is something I have written about at length in Your Newsletter Doesn’t Need More Subscribers. It Needs a Job.. AI is not coming for your inbox. A reader who subscribes has signed up to hear from you directly, regardless of how they found you in the first place.
Reviews still matter, on Amazon, on Goodreads, on StoryGraph. AI models are reading those reviews. The quality and volume of social proof has not stopped being relevant.
Author-controlled communities still hold weight. A Facebook reader group, a Discord, a Substack with engaged comments. Each of these is doing two things at once. They give you a direct relationship with readers, and they produce the kind of authentic, recent, on-topic content that AI models treat as signal.
And the unglamorous, structural SEO work I named at the top of this post still produces lift. Site speed, mobile responsiveness, internal linking, descriptive URLs, all of it still matters. AI search has not made traditional SEO go away. It has added a layer on top of it.
Where to put your effort
If I were sitting with an author who has three hours a month for website and SEO work, here is what I would tell them to prioritise, in order.
First, audit your About page through the lens I described. Is it making the claim that places you in the right corner of fiction for the right readers, or is it telling a personal story that assumes the reader already cares? Rewrite it so the first half functions as a source document and the second half does the personality work.
Second, audit your book pages for trope and structural language. If a reader asked an AI for "your" kind of book, would the AI know to suggest you? If the language on the page is too marketing-heavy, rewrite it to describe what the book is actually doing.
Third, add basic schema to your site. Most modern themes support this through plugins or built-in fields. If you do not know what you are doing, hire someone for an hour to set it up.
Everything beyond that is incremental. Useful, but not where the leverage is in 2026.
The strategic point
Author SEO has never been a one-and-done project. The reason it produces results across the careers I have worked on is that the fundamentals compound. A well-structured site gains visibility slowly, and that visibility persists. The shift toward AI search does not change that. It adds a question on top of it, which is whether your site is also legible to the models that are increasingly mediating reader discovery.
The work, in practice, is the same shape as it always was. Get specific. Use language that says what your book actually does. Maintain the basics. Let the structure earn its keep over time.
If you want help auditing your site through this lens, SEO Services is where we do that work. The audit pulls apart what is and is not functioning, including the new questions AI search has introduced.
