The Prompt I Run Before I Plan Any Content
Read Time: 4 Minutes
In today’s issue I’m giving you the one prompt I run before I plan any content, plus the audience map it built for my speaking channel.
If you post every week and get decent views but nobody books a call, this issue was made for you.
Aim at customers, not views
Caleb Ralston built the personal brands behind Alex Hormozi, Leila Hormozi, and Gary Vaynerchuk. He said something I keep coming back to. Most people chase views when customers are the thing that pays them.
Caleb on views versus customers. Open on Instagram
I built @prestonchinspeaks that way. I never tried to reach everyone. I picked one person and talked to him.
His name is Carlos.
A data scientist, 10 years in, mostly at big tech companies. He wants to move up by speaking better with leadership and with his team. English is his second language. He feels unsafe speaking off the cuff, so he scripts every presentation. He makes solid six figures and wants a leadership role, at his company or a new one, so interviews matter too.
- Carlos is a real student from The Impromptu Speakers Academy, the core offer in my coaching business. Anchor your own Core to a real customer or a real prospect conversation whenever you can.
- Pain matters more than demographics. People will self-select once you can describe their pain better than they can describe it themselves.
The prompt
Copy this. Paste it into ChatGPT or Claude.
I want a content strategy that pulls in the exact people who'd actually buy from me. I have a rough idea of who that is but not enough detail. Before you write anything, ask me the questions you need to pin down my one ideal viewer. What they do, what they're stuck on, and what I sell them. Ask a few at a time, and don't start until I say go. Then map the audience around that person in five layers. Core, Near, Mid, Wide, and Everyone, each one dropping a single detail so it gets broader. For each, tell me roughly how big it is and how close they are to buying. Then give me my content pillars and my next five videos. Three at Core, one at Near, one at Mid. Nothing wider than that.
How to run it
It asks you questions first. It will not start until you say go.
Answer like a person, not a marketer. Vague answers give you a persona that sounds like a textbook.
I did one thing that changed the output. I stopped answering from memory. I pasted in transcripts of real sales and coaching calls. I pasted in email replies from my subscribers. The replies helped most, because people describe their own problem in their own words. I got their language instead of mine.
Paste in whatever you have before you say go. A dozen real replies beat an hour of guessing.
The five layers
One person. Not a category.
Name a role and a narrowing detail together. A price point, a stage, a platform, a struggle. "Coaches" fails. Best case, you picture a real customer you could describe without thinking.
Same problem, one thing changed.
Still a group you could name out loud. The right people still spot themselves, so this layer reaches and converts.
Where new people find you.
This layer shows the platform there is a bigger pool next door to your Core, so it keeps sending fresh Core people back in.
The whole category.
Everyone is here, and so is everyone making content about it. Worth testing, because a hit out here feeds your Core. Most of the people it brings will never buy.
Big numbers, few buyers.
A hit out here still drags new Core people back toward the middle, so take a swing now and then. Read the number for what it is.
How I split the week
I post once a day. Out of every 7 videos:
So exact the right person thinks I made it for them. Fewer views, far more trust.
Close enough that the right people spot themselves, follow, and reply.
Mid, Wide, or Everyone. I rotate them and watch what comes back.
Five of the seven sit at Core and Near on purpose, because those two layers pay. The other two are tests. A big video out there is how new Core people find me, so I want them to travel.
Pick a split, run it for 2 or 3 batches of 10, then check which layer produced follows, DMs, and leads. Push the next batches toward that layer. Some businesses convert hardest at the centre. Some convert at Near. The numbers decide.
The map that built @prestonchinspeaks
Carlos. 10 years in, has to speak to leadership without a script, and freezes.
His job title barely matters. The pain is the whole definition.
Same professional, native speaker, same fear of the unscripted moment.
Same problem, one less thing making it worse.
Any professional who wants to present better at work.
Prepared talks, decks, meetings. Every communications coach already lives here.
Anyone who wants to be a better public speaker.
Students, creators, founders, wedding toasts.
Anyone who wants to feel more confident.
Self improvement. Huge, and nobody here buys a speaking program.
116 people on my newsletter and in my community told me why they wanted to speak better, and the loudest answer, confidence in public speaking, is a Mid answer that every communications coach already owns. The two dozen people who paid for the first cohort of the Impromptu Speakers Academy described something far narrower: making a point on the spot, in front of senior leaders, without rambling. So I stopped making content about public speaking and started making it about impromptu speaking, a category nobody owned and the exact thing Carlos could not do.
That is the payoff from this exercise. A category small enough to own.
The map I run now
I ran the same exercise for @prestonchinbuilds, my second channel.
Making general AI content would have been the easy move, and that lane is packed. Most accounts post the same tool roundups and the same trends, and none of them say who the content is for.
So I picked a Core instead. Creator-educators like me, people who sell a program and want short-form video to bring in leads. Then I picked one angle to own. I show the real systems I have built that run my own coaching business with AI. I skip the hype and the trend chasing. If I have not shipped it and used it myself, it does not go on the channel.
Coaches and consultants sitting on hours of their own recorded expertise.
You, minus the ability to build it.
Coaches and consultants who want content to bring in leads, backlog or not.
Same buyer, same offer, no pile of recordings to start from.
Creators who want more views without sounding like everyone else.
No offer needed. Just an audience and a voice they are losing. This is my reach slot.
Small business owners who want AI inside their operations.
Content becomes one use case among many. Marketing automation, company brains.
Anyone trying to use AI to get more done.
The junk-food ring. Big numbers, no buyers.
Look at how little sits between Core and Near. One detail. That is why "coaches and consultants" on its own fails as a Core answer.
Where the ideas come from
People always ask me this next: nobody makes content for my exact Core, so who do I study?
Take topics from Core and Near. Mid at a push. Borrow topics from the outer layers and you drift broad without noticing. If no creator sits in your Core, that is your opening. Grab a Near creator's topic and re-aim it at your Core.
Take craft from anywhere. Formats, hooks, story shapes, pacing, editing, thumbnails. A cooking creator's format will carry your topic just fine.
Free resources
Once you have your map and your pillars, the next problem is ideas that feel flat.
- The 3D Idea Matrix takes one pillar and multiplies it by 50 angles and 12 formats, 10 ideas at a time.
- The Content OS is the database and the agent I run everything through once the ideas exist.
The five layers and the video mix come from Kane Kallaway's Bullseye Method, which he built out of making 367 short-form videos. Define, hit, expand. I renamed the layers and wired the whole thing into the prompt above.
Want me to build a content system or an agent for your business? Tell me what you need and I'll get back to you personally.
Preston
Get a new build-with-AI guide every week
Weekly playbooks on prompting, agents, and content systems — free.
You're in. First email lands shortly.