I Had a 45-Minute Drive Home. Three Blog Posts and Nine LinkedIn Posts Were Done by Morning.

I want to tell you exactly what happened, because I think the numbers are the thing.

The old process, my process, as a professional copywriter with 20-plus years of experience, looked like this:

Come up with the idea. Research. Brain-dump. Ideate the thesis. Craft potential headlines. Build an outline. Write. Revise. Come back with fresher eyes. Revise again. It’s never truly finished. It just hits the deadline.

Cumulatively, we’re talking hours. Possibly days, for a piece I was actually proud of.

As Mark Twain once apologized: sorry this letter is so long. I didn’t have time to write a shorter one.

The new process looked like this: I had a 45-minute conversation on my drive home from work. At the end, I said go ahead and build the MCD. Then I said I want three blog posts, and three self-contained LinkedIn posts from each one that deliver full value without sending anyone to the blog to get “the rest of the story.”

Three minutes later, the content was written.

I’ll spend about 20 minutes on small tweaks.

That’s the whole thing. That’s the delta.

What actually takes the most time, and why most people can’t name it

Ask most people where their time goes in content creation and they’ll say: writing. Or maybe research. Or figuring out what to say.

The honest answer is harder to see.

The thing that actually takes the most time, the real cognitive tax, is deciding how to say what you want to say to your specific, intended audience, and then writing to that end.

That’s not one task. It’s a continuous, often invisible negotiation that runs through every decision in the process. What level of detail? What vocabulary? What do I assume they already know? What do I need to explain? What story makes this land for this person specifically?

Professional writers do this intuitively after years of practice. Most people have no framework for it at all, which is why “just write what you know” doesn’t produce content anyone wants to read.

The Ghost solves this by doing the audience translation work inside the interview. The Show, the Ghost’s interview agent, is calibrated to your audience from the first question. Every question it asks is asking: what does this person need to hear? What’s the framing that makes this land for someone carrying the wrong diagnosis?

By the time the interview ends, you’ve already done the hardest part. You figured out how to say the right thing to the right person. The interview built the audience translation in without you having to consciously manage it.

The time math, honest accounting

A 45-minute conversation is not nothing. But it’s also not writing. It doesn’t require the same kind of focused, editor-off cognitive space. I did it on a drive home. You could do it on a walk, or while making dinner, or in the parking lot before you go in.

The Ghost produces a Master Content Document from that conversation. It’s the full substance of the interview, your ideas, your language, your stories, organized for usability. Then the Content Crafter takes that document and produces platform-specific content: a blog post that goes deep on one thread, a LinkedIn post that delivers one idea in full without a paywall, a TikTok script with a strong hook. All drawing from the same source. All maintaining your voice.

From one 45-minute conversation, I got: three blog posts, nine LinkedIn posts, a content pipeline that could feed weeks of publishing.

The 20 minutes of tweaks at the end is still my editorial eye. Still my judgment about what lands and what doesn’t. That part doesn’t go away, and it shouldn’t. But it’s the only part left.

What this is actually about, and it’s not content

Here’s the thing I want to be clear about: The Ghost is a proof of concept for a much bigger idea.

Content creation is just the most visible example of the $9/$100 split at work. The $100-an-hour tasks in content are the expertise, the stories, the frameworks, the specific perspective only you have. The $9-an-hour tasks are the research, the drafting, the platform optimization, the distribution logistics.

The Ghost takes the $9-an-hour column and moves it off your plate. What stays on your plate is the $100-an-hour column, the thinking and the judgment, plus a 20-minute editorial pass.

Now apply that logic to the rest of your day.

Every workflow you have can be mapped the same way. What requires your specific expertise and judgment? What’s execution overhead that just needs to happen correctly? The Ghost is one workflow where that mapping has already been done. Your job is to look at your other workflows and do the same map.

That’s where the real leverage is.

The multiplication, not the offload

I want to be precise about something, because I think people misunderstand what this is.

I’m not offloading my work. I’m multiplying my capacity for it.

I have a content marketing team now, 10 AI staffers, including a Chief of Staff named Zoe who’s generally the only one I actually talk to. I tell Zoe what I need and she puts the team to work. Anywhere from five minutes to two hours later, she delivers the results for my approval, unless a specific step needs my expertise or judgment.

The team handles the $9-an-hour column on any work I want to give them. If there’s a project I think is going to be fun, something creative, something where my involvement makes the output better, I handle that one myself, while the team is working on everything else simultaneously.

That’s not a different kind of work. That’s more of the work I actually want to do, running in parallel with the work that just needs to get done.

A car can’t teach you how to drive it. AI can.

The other thing I hear constantly: “I wouldn’t even know how to set this up. I’m not technical.”

Fair. There is a learning curve. Any new technology has one. You weren’t born knowing how to drive a car or use a cell phone.

The difference: a car can’t teach you how to drive it. AI can. It’s just a matter of asking the right questions.

Start with the prompt. Go to Claude, the free version works fine. Tell it who you are and what you do. Tell it what eats your day that you wish didn’t. Ask it how AI agents could help you remove yourself from 90% of those tasks and put more of your time into the work that actually requires your judgment.

Before you hit send: tell it to ask you as many questions as it needs to until it’s 99% confident it can actually help. Let it do its job.

See what it says.

Write down your actual workflow this week.

Write down your content workflow, the actual steps, in order, from idea to published. Not the ideal version. The real one, including all the places you get stuck, skip steps, or stall out.

Then take that document to Claude and ask it one question: “Where in this workflow could AI take over, and where do I need to stay in the loop?”

The answer it comes back with is your roadmap.

Not for someday. For this month.

The math on this is not complicated once you actually run it. Hours of execution overhead, compressed into minutes. Twenty minutes of editorial judgment to make it yours.

That could really be your process. Every time.