I was browsing the settings menu in Claude Desktop. Not hunting. Just opening drawers, like you would in a kitchen looking for scissors.
Under a customize option, tucked into a plugin panel, there was a Marketing section. The other plugins I had already set up sat out in the open. This one was behind a drawer. Inside: brand review, campaign planning, competitive briefs, content creation, email sequences, performance reports, and at the bottom, seo-audit.
I have an MBA in marketing I never used. I went to business school and then went back to writing software. The degree has sat on my résumé for fifteen years, quiet, like an unused language. But when I saw the word, something in me leaned toward it. I installed the SEO skill.
I ran it on my own website.
The report came back in ninety seconds.
Of 979 pages, Google had indexed 94. That is 9.6 percent. 1,310 of my posts had zero incoming links from anywhere else on my site. 1,293 posts had no focus keyword set. 419 of my external links were broken.
Twelve hundred posts. Twenty years of writing. Ninety-four of them showing up in search.
Here is what I thought my website was. Twelve hundred posts. Four books. A subscriber base. A clean theme. The home of a working writer on the internet. If you arrived at it, you would find the work.
Here is what the numbers said it was. A collection of files. The essays did not link to each other. The book pages did not link to the essays that had led to the books. The old posts did not know the new posts existed. Google had crawled the pages and given up on most of them, because nothing told it which ones mattered.
I had built it myself. Not by choice. By omission. By uploading a post and moving to the next one, because that is what writers are told to do, and that is what I had always done, without anyone telling me that posting was not the same as publishing on a site that could be found.
I turned fifty in December.
The party happened. After the guests left and the gifts were opened and the rest of the family went to bed, what I did was start asking a question I had been pushing away for a few years, which was, what do I want to be known for.
I have been writing since Sulekha discussion boards in the late nineties. Writing has been the one constant across everything else. The software career, the years at home with the twins, the break, the return. Books came. Essays came. Readers came. A feature documentary came, not mine, but built on a life I had been living openly for years. By any visible measure, the writing was working.
What it was not doing was adding up.
I did not mean sales. I meant the other kind of adding up. The one where a reader finds one essay and follows a thread through the rest. Where a body of work signals to someone new that here is a writer who has been thinking about something for a long time, and here is where her thinking has been. I had twelve hundred posts and four books and essays in anthologies and a documentary on the festival circuit. And if a stranger arrived at my website, I was not sure they would find the throughline. I was not sure there was one to find.
Fifty is a year of asking what your work is for. I did not want to be known for being prolific. Prolific is not a legacy. I wanted to be known for having said something true about being a Tamil American woman raising children who do not share my blood, about identity and belonging and what gets passed down, about the essay as a form that can hold complexity without resolving it. I wanted my words to leave a mark.
Then I started reverse engineering.
How does a writer become known. What makes a search engine surface her when someone types in a question she has answers for. What makes an AI product, when asked for a credible voice on adoption or identity or the Tamil diaspora, mention her. These are not the questions I was trained to ask as a writer. They are the questions I was being asked to answer by the landscape.
All three questions had the same answer underneath them: discoverability. Not marketing. Not selling. Discoverability. Would a reader wanting to find the work, find it?
The SEO plugin was how that philosophical question turned into a practical one.
I had already been working on the website. A redesign plan had been in motion since March. New home page. New books page. A Film page for the documentary. An Essays page to replace the old Blog. The tagline had been changed from Author. Parent. to Essayist. Author. Memoirist. The structural work was underway.
But ten days before I ran the SEO audit, a content migration had gone badly wrong on the same site. What mattered was that I had just watched my own site misbehave in public. I wanted to know what else I did not know about it.
So I ran the audit, and the numbers came back, and I stopped working on anything else for the rest of the morning.
The structural problem was not that individual essays were badly written or poorly tagged. The structural problem was that the essays did not know about each other. There was no topic cluster architecture. No pillar pages. No hub that said this writer has thought about transracial adoption for fifteen years and here is the map of her thinking. Each essay sat in a chronological pile, and Google, looking at the pile, saw no reason to promote any one essay over another. There was no signal of authority because there was no structure that claimed authority.
And the site was leaking. The homepage meta description referenced a project I had been careful not to discuss in public. The public-facing infrastructure was saying things I had never authorized it to say. I fixed that first.
Then I started on the pillars.
Pillar pages are a technical SEO concept, but they are also what a serious writer is doing when she puts together a book. You pick a topic. You show where your thinking on it has been. You link the essays together so a reader arriving at any one of them can find the rest. You tell the machine and you tell the human the same message: this is a body of work, not a pile of posts.
I had not thought of my writing as a body of work. I had thought of it as what I had written this week, and the week before, and the week before that. The SEO audit was asking me to think of it as architecture. The reverse engineering I had been doing philosophically turned out to be, at the practical level, a request that I finally organize twenty years of writing into something that could be seen whole.
The infrastructure followed from that. A biweekly SEO audit, scheduled. Four custom code snippets on WordPress doing work the existing tools could not. An llms.txt file at lgiyer.com/llms.txt, because the crawlers that matter are no longer only Google or Bing. Large language models may eventually read my site through this file, and I would rather be early than absent. If they do read it, what I put there will shape whether I am mentioned when someone asks them for a voice on the topics I write about. If they do not, I have lost nothing but the hour it took to write.
That is what the philosophical question looked like six weeks later. Scheduled tasks. A backlog of things to fix. A git repo. A writer running a small infrastructure system between dinner and dawn, because being findable is a form of labor, and the writers who are not doing it are going to be the writers who are not found.
The other pieces are still the writing. Still the essays on Monday morning. Still the book that is out to agents and the book that is coming from a publisher and the book in draft and the one I have not started. Being findable is not the work. It is what lets the work reach the reader.
At fifty, that is the version of legacy I am prepared to build for. Not viral. Not a bestseller list. A body of work that the person who would want it can arrive at, from any direction, and follow.
The AI Journey
- The Morning I Downloaded VS Code — The morning I opened the terminal I had walked away from twice.
- Two Hours on a Monday Morning — A publishing pipeline, five weeks of failure, and the fifteen-minute replacement.
- Nine Point Six Percent — The SEO audit that asked me what twenty years of writing were for.
- The Librarian and the Historian — Reading across fifteen years of correspondence for a memoir.
- What the Dashboard Doesn’t Count — The cost of the migration that went wrong.
- Saturday Morning — The practice I have concluded for myself.
Leave a Reply