Last week I invoiced a client $520 for a 2,600-word article on building effective microlearning modules for corporate training. I charge $0.20 per word, and most client articles run between 2,000-2,800 words. Total time from brief to final draft: just under an hour.
That's $500 an hour.
I'm not telling you this to brag, well maybe just a little. I'm telling you because if you've spent years building expertise in your field (whether that's HR, finance, healthcare, operations, whatever) you're probably undervaluing what you know. And you're definitely not using AI the way you could be.
AI doesn't make writing easier for everyone. It makes writing faster for people who already know what good looks like.
The $500/hour isn't Claude's rate. It's mine. The AI is just the infrastructure that lets me deliver expert work in a fraction of the time.
The One-Time Setup That Makes This Possible
I don't sit down and prompt Claude from scratch every time I get a client brief. That would be insane and wildly inefficient.
Instead, I spent about 6-8 hours building a system (once) that I now reuse for every piece of writing work I do. The system has four parts:
- A client brief template that captures everything I need to know upfront
- A style guide framework that translates a client's voice into something Claude can follow
- A quality checklist that ensures nothing leaves my desk unless it meets professional standards
- Research and editorial standards that keep the work substantiated and credible
Once these are built, they're reusable. I refine them occasionally, but the core infrastructure stays the same.
This is the setup tax. You pay it once, and then you get the benefits forever.
How the Flywheel Actually Works
Let me walk you through that $520 instructional design article so you can see what this looks like in practice.
Step 1: The Brief Comes In (5 minutes)
Client sends me a brief. It's comprehensive: includes target audience, user intent, angle, word count, internal links, expert quotes, competitive articles to reference, SEO keywords.
I don't need to guess what they want. It's all there.
Step 2: Load the Brief Into My Claude Project (2 minutes)
Here's where the infrastructure pays off.
I have a Claude Project set up with all my reusable documents already loaded:
- My editorial standards doc
- My quality checklist
- Research standards
- General style frameworks
These live in the project permanently. I never have to re-upload them.
When a new brief comes in, I just:
- Open a new chat in that project
- Upload the client brief
- Upload their specific style guide (if they have one)
That's it. Claude already has access to all my standards and frameworks because they're in the project knowledge. I'm just adding the client-specific stuff.
Then I say: "You're writing this article. Follow the brief exactly. Use the client's style guide for voice and formatting. Meet all the editorial standards in the project knowledge. I'll edit the output before it ships."
No elaborate prompting. No re-explaining my standards every time. The infrastructure is already there.
Step 3: Claude Writes the First Draft (Instant)
Claude reads everything, processes the brief, and writes a 2,600-word draft on microlearning design principles. Takes maybe 60 seconds.
Is it perfect? No. But it's structurally sound, on-brand, and addresses user intent. The hard work (research, quotes, formatting, internal links) is done.
Step 4: I Edit and Shape (30-45 minutes)
This is where my 20 years of experience matters.
I'm not rewriting from scratch. I'm:
- Tightening sentences that are too verbose
- Cutting jargon that crept in
- Adjusting tone where it doesn't quite land
- Fact-checking claims against the research
- Making sure the expert quotes are woven in naturally, not dumped in blocks
- Ensuring the piece actually solves the reader's problem
I know what good L&D writing looks like. I know when something sounds like generic AI slop versus useful, human-centered advice. I can spot when a concept isn't explained clearly or when an example doesn't work for the target audience.
That judgment? That's what clients pay for. Claude can't do that part.
Step 5: Final Quality Check (10 minutes)
I run the article through my quality checklist:
- Does it answer user intent?
- Is it the right reading level for the audience?
- Are all facts substantiated?
- Is the tone consistent?
- Are there any grammar/spelling errors?
- Does the intro hook and the conclusion land?
If something's off, I fix it. If it's good, I ship it.
Total time: Under an hour. Payment: $500-560. Rate: $500+/hour.
Why This Works for Me (And Might Not Work for You)
Let me be very clear about something: This only works because I'm an expert.
I've spent 20+ years in learning and development, HR, and organizational change. I've written thousands of pieces of content. I've worked with dozens of clients across industries. I know what works and what doesn't.
That expertise shows up in three critical places:
1. Setup. I know what information to gather upfront. My brief template isn't generic. It's designed to capture the exact details that will make the writing process smooth. If you don't know what questions to ask, you can't build good infrastructure.
2. Editing. I can read Claude's output and immediately see what's wrong. A sentence that's too long. A concept that's explained poorly. A tone that's off-brand. If you don't have taste, you can't fix bad writing (even if AI generated it).
3. Quality control. I know what professional-grade work looks like because I've been producing it for decades. My checklist isn't aspirational. It's based on real standards I've met over and over. If you don't know what "good" is, you can't hold AI (or yourself) to it.
This is why junior writers can't do what I do, even with access to the same tools. They don't have the judgment yet. They don't know when Claude's bullshitting or when a piece of advice sounds smart but won't actually work in practice.
If you're early in your career, AI will help you write faster. But it won't make you an expert.
If you've been in your field for 10+ years, AI will let you monetize your expertise at a completely different level.
What This Really Means: Knowledge Is Valuable
Here's the uncomfortable truth: A lot of experienced professionals don't realize how valuable their knowledge is.
You've been in your industry for 15 years. You've seen what works and what doesn't. You can spot a bad strategy from a mile away. You know the nuances that textbooks don't teach.
But because it feels easy to you, you assume it's easy for everyone. So you underprice yourself. Or you don't sell your expertise at all.
AI doesn't change the fact that expertise is valuable. It just makes it easier to deliver that expertise at scale.
If you're a senior accountant, you could use this system to write financial explainers for clients.
If you're a healthcare administrator, you could write guides on compliance or operations.
If you're an instructional designer, you could write about learning theory, course design, or elearning development (or even build and design actual elearning courses faster).
The setup is the same. The infrastructure is reusable. The only variable is: Do you actually know what you're talking about?
The Templates
Here are the four documents I actually use, sanitized and turned into blank templates you can adapt for your own work:
Free $500 an Hour Writing TemplatesYou'll get:
- Client brief template
- Style guide framework
- Quality checklist
- Editorial and research standards
These aren't magic. They're just structure. But structure is what separates people who use AI effectively from people who generate slop and hope it works.
Use them. Adapt them. Build your own system.
A Few Caveats Before You Run Off and Try This
This isn't passive income. I still trade time for money. It's just that my time is now worth more because I can deliver faster without sacrificing quality.
Clients need to be organized. If a client can't provide a decent brief, this system breaks down. I spend a lot of time upfront training clients on what I need. That's part of the work.
You still need to edit. If you ship Claude's raw output, you will get caught. It will sound generic. It will miss nuance. Your reputation will tank. Don't skip the editing phase.
AI makes mistakes. Claude hallucinates facts. It invents sources. It gets concepts wrong. If you don't know the subject well enough to catch these errors, you will publish incorrect information. That's on you, not the AI.
What I'd Do Differently If I Were Starting Today
If I didn't have 20 years of expertise and wanted to build toward this kind of work, here's what I'd focus on:
1. Pick a niche and go deep. Don't try to be a generalist. Become the person who knows one thing better than almost anyone. That's where the premium rates live.
2. Build taste by studying great work. Read the best writing in your field. Analyze why it's good. Learn to spot the difference between "this sounds smart" and "this is actually useful."
3. Start building your infrastructure early. Even if you're junior, create templates. Write down what good looks like. Build checklists. This will make you faster and better over time.
4. Charge what you're worth. Seriously. If you're experienced, your knowledge has value. Don't compete with people who are just starting out. Compete with people who've been doing this as long as you have.
Final Thought: Sell Your Knowledge
If you've been working in your field for a decade or more, you have expertise that other people need. AI just makes it easier to package and deliver that expertise.
But only if you use it as infrastructure, not as a replacement for judgment.
The $500/hour isn't about the tool. It's about knowing what to do with it.