Selectable - Part 2
What happens when a reader asks AI for a newsletter like yours — and AI can't describe what you do.
Last week, I talked about being selectable for AI for book authors. You can find that here.
We talked about genre, tropes, and comp titles. Things books need to convince readers to make the purchase. The external representations of the internal promise an author makes to their reader. How AI needs these to be clear and consistent to select the book.
But that’s not who you are.
You’re a Substacker. You write a newsletter about money, life, art, how to make money on Substack, how to grow old, health matters, and the whole wide array of topics which reside here. So how can you make AI select you when someone asks about your topic?
That’s what we’re going to cover today.
You’re probably not invisible anymore
If you’ve read Findable — the first post in this series — and done the Google Search Console setup, you’ve likely fixed the biggest gap. Google can see you. Which means AI can find you.
But findable and selectable are two very different things.
Findable means the box exists and Google can open it.
Selectable means AI knows what’s inside — and can match it to a reader’s request.
A newsletter that’s indexed but vague is like a shop with a door but no sign. You can walk in. You have no idea what’s being sold.
This is the undescribable problem. It happens when AI found you but couldn’t figure out what you do. And this is more common than being invisible.
Every newsletter makes a promise.
Book authors make an internal promise — the one they carry in their heads while they’re writing, the one they want to deliver to their reader. Then they translate it into genre labels, tropes, and comp titles. It is this translation that helps readers pick their work.
Substackers make the same internal promise. Most just haven’t translated it yet.
It might be the insider promise — I’ll pull back the curtain on what’s really happening, not the version everyone performs online. Or the community promise — I’ll make you feel like you’ve found your people, the ones who think the way you do. Or the growth promise — every post moves you one step closer to the thing you’re trying to build.
Most writers carry one of these — sometimes all three — without ever writing them down. They just write toward them, week after week.
You know who you’re writing for. You know what you want them to feel when they read your newsletter. You know what you’re not — you’re not just “another writing newsletter” or “thoughts on life.” You have a specific thing you do for a specific kind of reader.
The question is whether AI can see that.
The three places the promise has to live
Your tagline
The first thing AI reads — and the first thing a new reader sees. It has one job: tell them immediately whether this newsletter is for them.
“Thoughts on writing and creativity” is a promise so broad it describes half of Substack. “Weekly notes for fiction writers learning to use AI as a creative partner — not a ghostwriter” is a promise AI can actually work with.
My tagline used to read “Exploring the edges of Writing, Tech, and Imagination.” And while that’s still true, it could describe a science blogger, a tech journalist, or someone who takes apart toasters for fun. I rewrote it to “exploring AI in the systems writers use.” Now there’s no guessing. Writers who are trying to figure out where AI fits in their work know immediately — this newsletter is for them. And AI knows who to recommend me to.
About You
Substack gives you enough room for a short paragraph to describe yourself. But don’t underestimate it. This is where curious readers go to find out about you. And where AI goes to understand the full picture.
It has one job: tell anyone who lands there — human or AI — exactly who you are, who you write for, and what they get that they can’t get anywhere else.
If it says “Hi, I’m [name] and I love writing and exploring ideas,” AI has to guess if you’re giving writing advice or exploring science, self-help, money. And guessing is where the mismatches happen — wrong readers, wrong recommendations, wrong expectations.
Mine reads: “I’m a fantasy-fiction author and recovering IT nerd. I help established authors become AI-visible — without compromising voice, craft, or creative control. Expect systems talk, story talk, and the occasional witchy metaphor.”
Three sentences. You know exactly who I am, what I do, and what I write about. A reader looking for hustle-culture content marketing advice will self-select out immediately. A writer who wants sane, grounded AI systems thinking will feel like they’ve found their people.
That’s the job. Not impressive. Not clever. Just clear.
Your recent posts
AI doesn’t just read what you say you do. It reads what you actually publish and builds a pattern. If your tagline says one thing and your last ten posts are about something else entirely, AI will trust the posts.
The underlying thread is the thing to pay attention to. Not whether every post covers the same topic — it shouldn’t. But whether a reader who lands on any post from the last few months could recognize the same writer, the same angle, the same reader being served.
Look at my last four posts. AI is becoming the discovery layer for books and newsletters — and most authors don’t know it yet. Claude wrote me a publishable Note and I rewrote the whole thing. My own newsletter was invisible to AI, and one morning fixed it. Genre, tropes, and comp titles are how AI figures out which book to put in front of which reader.
Four different topics. One unmistakable thread: AI belongs in your systems, not in your sentences. Every post circles back to it from a different angle.
That’s what AI is looking for when it reads your archive. Not repetition. Recognition.
One newsletter. One story.
The most important thing about all three parts? They have to say the same thing.
This is the gap for most Substackers. You can have a sharp tagline, a solid About paragraph, and a consistent publishing record. But if those three things are pointing in different directions — different audiences, different topics, different promises — AI has to guess.
Book authors have comp titles to help with this — three comparable books that tell a reader “if those worked for you, this probably will too.” Substackers don’t have that tool. Readers don’t say “I want a newsletter like X.” They say “I want something about Y for people like me.”
Which means your three areas — tagline, About paragraph, recent posts — are doing the job comp titles do for authors. They’re the items AI uses to decide whether you match the request.
The more they agree, the easier the match.---
Don’t be fooled — the work takes awhile. For me, I had to figure out what I actually wanted to write about first. That took experimentation. Once I figured that out, I had to get clear what I could help writers with when it came AI.
Most of what I see out there is about AI and writing — using it to draft, to generate, to produce. To me, that’s like asking Word to add up a column of numbers. Technically possible. Not what it’s built for.
The revelation came when I asked AI to help me design a system I could actually follow. Something that would make my writing life more consistent, less chaotic. That’s when it clicked. It can design systems that work for me. I don’t have to keep failing in someone else’s system. That’s what I help writers find. AI inside the systems they use.
A three-question audit you can run right now
Read your tagline, your About paragraph, and your three most recent posts. Then ask:
1. Could a stranger describe what this newsletter does in one sentence — without reading more than the tagline?
2. Does your About name your specific reader — not “anyone interested in X,” but the actual person you have in mind?
3. Do your recent posts reinforce the same promise, or have you drifted?
If you can answer yes to all three, AI can describe you. And a reader asking the right question will find you.
Unexpected bonus
Here’s what surprised me most about doing this work.
The clarity you create for AI? It does the same thing for your Substack newsletter.
When your tagline, your About paragraph, and your recent posts all say the same thing — something shifts. The readers who were meant to find you, find you faster. The ones who weren’t, self-select out.
Your comments start coming from people who get it. Your replies feel like a conversation with people who were looking for you.
That’s the transformation hiding inside what looks like a technical exercise. You’re not just optimising for AI. You’re finally saying out loud — clearly, consistently, publicly — who this newsletter is for and what it does.
You know the posts about how to grow on Substack? This is the foundation underneath them.
Next up: Trustable — what happens when AI can find you, describe you, and has to decide whether to trust you so it can recommend you.