50 Orders from ChatGPT
The new question isn’t “can you be found?” It’s “can you be recommended?”
Over Christmas, I fell into a business rabbit hole that had nothing to do with books.
A company received 50 orders from ChatGPT.
Not clicks.
Not traffic.
Orders—placed inside the conversation.
I sat there with actual goosebumps, because once you see that shift… you can’t unsee it.
Discovery isn’t starting with browsing anymore.
It’s starting with asking.
And here’s why that matters for us:
while the conversation keeps circling AI slop and whether AI is “writing books”…
the bigger shift is happening quietly underneath:
Discovery is becoming conversational.
And when discovery happens inside a conversation, the question isn’t just,
“Can you be found?”
It’s: Can you be recommended?
At my desk
I did what every rational person does after reading something mildly terrifying.
I swore! Like a trooper.
Because I suddenly realized: if AI discovery is moving into conversation, then the “unit” we’re optimizing isn’t a webpage.
It’s not even SEO.
It’s not “content.”
It’s… a sentence.
A sentence that tells a system (and a reader) who you are and what to do with you.
So I went looking for the sentence I thought I had—
and discovered I had three half-sentences pretending to be one.
The villain
The villain isn’t AI.
It’s the myth we are still living under:
“If people can find me, I’ll be fine.”
That was true when discovery started with browsing:
Search → scroll → click → compare → decide.
Now it’s becoming:
Ask → answer → decide.
So being findable? That’s the door that lets you in.
The real game is whether you’re recommendable:
easy to place, easy to match, easy to bet on.
(Substack-only writers: you’re not exempt. Your Substack is your footprint—profile, pinned post, repeated language. If someone were to ask AI for a newsletter on your topic… can it find you and name you?)
The annoying truth
This is what clicked for me at my desk:
AI is basically a matchmaking engine with no patience.
If it can’t place you quickly, it moves on.
If it guesses wrong, it creates a mismatch—wrong expectations, wrong readers, confusion.
So when I say “I went hunting for my sentence,” this is why.
Because I’m building a business around helping authors become easier to recommend—
and I’m not doing that from a foggy foundation.
. . . . .
The two-line spine
This week, I’ve been tightening clarity on the consulting side of my ecosystem—because this is the work I teach.
So I built it in two sizes—headline and supporting sentence—and I’m installing it everywhere:
I help established authors become AI-visible—without compromising voice, craft, or creative control.
And the longer version when there’s room to explain:
I help working authors put AI in the systems around their books—not in their storytelling—so their work becomes findable, selectable, and trustable wherever readers ask for recommendations.
Same promise. Two sizes.
One for instant recognition. One for trust.
Not “AI writes your book.”
Not “AI hacks the algorithm.”
AI as an assistant and systems partner—so your work becomes easier to place, easier to match, and easier to choose.
What I mean by “one sentence”
I don’t mean a clever tagline.
I mean a sentence that does three jobs in public: who it’s for, what changes, and what it’s not.
Because that’s what helps a reader (or an AI) go:
“Ah. This is what I want.”
. . . . .
Five places I’m updating
1) Bio / positioning line
First fix: I put my two-sentence positioning stack where people actually meet me—my Substack profile and my LinkedIn profile.
If someone lands on either one, they should immediately see:
what I do
who it’s for
and what I’m not doing (AI in the system, not in the storytelling)
If you only read those two spots, you should still be able to say:
“Oh—got it. I know what she does. And I know my storytelling stays mine.”
2) Pinned post
Next fix: I’m writing a pinned post that functions like a “Start Here” for new people—without requiring them to go hunting.
It’ll live on Substack, and a version of it will live on my website.
The goal is simple: if someone finds me through a recommendation, a share, or a random Notes rabbit hole… they can land, read one post, and immediately know:
what I’m about
who I help
and where to start
3) Recurring language inside Notes + posts
Then I’m reinforcing it the only way that actually sticks: repetition in public.
Next fix: I’m getting intentional about the phrases I repeat—because repetition is how people learn what you mean.
So in my Notes and posts, you’ll keep seeing the same anchors show up in plain language:
assistant and systems partner—not a ghostwriter
AI in the system, not in the storytelling
findable / selectable / trustable
Not as slogans. More like signposts.
Because one good bio isn’t enough. The message has to hold up across a few weeks of real writing.
4) Proof + trust signals
Final fix: I’m backing up my claims with visible proof—so a new reader doesn’t have to take my word for it.
So I’m adding a few concrete trust signals across my Substack and website:
a clean “what we cover” list (so people know what they’re actually signing up for)
before/after examples (bios, blurbs, book descriptions, series order pages)
“this is for you if…” / “not for you if…” lines (so the right people self-select)
and a plain-English explanation of how AI is being used (systems, analysis, structure—not writing anyone’s voice)
In my view, trust gets built the same way every time:
what you say about your work matches what people can see.
5) Consulting offer page (in progress)
Next fix (still in progress): I’m building the page for my consulting offer using the same two-sentence spine—same language, same terms, same boundaries.
Because if my Substack and LinkedIn say one thing…
and my offer page says another…
I’m basically asking readers (and AI) to guess which version is true.
So I’m setting it up using the structure that works:
what the offer actually does (the outcome)
what it doesn’t do (the boundary)
and what you can realistically expect to walk away with
. . . . .
Try it
If you want to test your clarity, write one sentence a stranger could repeat accurately. Bonus points if an AI could, too.
It should cover who it’s for, what changes, and what it’s not.
1. The offer sentence
“I help ___ do ___ by ___ — without ___.”
2. The reader-fit sentence
“It’s for readers who want ___ (feeling/vibe) with ___ (tropes/themes) in a ___ (genre lane).”3. The boundary sentence
“This is not ___. This is ___.”
If your first draft sounds like generic internet language, don’t panic.
That’s just your cue to get more specific.
Want help?
If you’re stuck defining:
your bio
your offer
your blurb
your book description
your genre/vibe sentence
Leave me a comment with “clarity check” and I’ll point you to the single highest-leverage fix—the one that makes everything else easier to write.
Next up: I’ll show you how I’m turning these sentences into a repeatable public footprint—the part that makes “posting into the void” start compounding.
Helping established authors become AI-visible—without compromising voice, craft, or creative control.
Website: AIMagicforAuthors.com
E-Mail: info@aimagicforauthors.com