At the end of 2024, I was tired of being a generic content writer.
I worked 60+ hours most weeks, yet still struggled to make enough money to make ends meet.
Something needed to change.
So I narrowed my focus and created a ghostwriting offer specifically for founders of companies that trained people about AI.
Now I just needed someone to buy it.
Fortunately, I already knew a few people who would be a great fit.
A couple of years earlier, I worked as a machine learning engineer, which helped me rub shoulders with some super smart people.
During that time, I met someone who would eventually become my first ghostwriting client.
At the start of 2023, he reached out for advice on building his online presence.
I’d already spent years writing online by this point, so I gave him everything I knew.
We talked about content.
Positioning.
Building an audience.
We even built a project together that he wrote about and shared on Medium.
The post blew up.
It was his first taste of virality.
But after that engagement, we both went our separate ways.
Fast-forward to December 2024…
We crossed paths again.
By this time, his LinkedIn following had shot up to around 52,000 followers.
Clearly, he’d done pretty damn well without me.
But despite building an audience most people would kill for, he needed help taking things further.
And this time, instead of giving him advice for free…
He became my first ghostwriting client.
Over the following 21 months, we published hundreds of posts, experimented with different ideas, and learned a hell of a lot about growing an AI engineer’s personal brand on LinkedIn.
Today, he has more than 112,000 LinkedIn followers.
That’s over 60,000 additional followers since we started working together.
In the remainder of this article, I’ll break down exactly what we did, what worked, and what I’d do differently if I were starting again today.
The client and where we started
By the time we started working together, my client had already cracked LinkedIn growth.
But it came at a major cost…
Time.
Creating content was taking him up to 10-20 hours a week.
This meant he had less time to focus on building, which was what he loved.
Building also helped him stand out from other creators writing low-level, generic content at high frequency.
Most people were just paraphrasing what others were saying.
You could tell he knew what he was talking about.
He built and showed you how to draw your own conclusions.
This was a big reason he had grown so quickly.
After I gave him some advice on building his online presence in 2023, he went all in.
He:
Created free courses to help machine learning engineers improve their skills.
Shared what he was learning.
Documented what he was building.
Published content 7 days a week.
It worked.
By December 2024, he had around 52,000 LinkedIn followers, and his posts were regularly reaching large audiences.
So there wasn’t some broken content strategy for me to come in and fix.
If anything, that would’ve been stupid.
My job was to preserve what was already working while removing him from as much of the writing process as possible.
He wanted to get back to coding.
I wanted to take the responsibility of creating content off his plate without making the content worse.
Fortunately, we already had a lot of trust.
We’d known each other since my machine learning engineering days.
He’d experienced my writing firsthand.
We’d built something together.
So unlike most new client engagements, I didn’t have to spend much time proving anything.
The bigger challenge was extracting the complex AI concepts from him and explaining them in a way that’s consumable and exciting for his audience.
The strategy behind 60,000+ new followers
This client is different from anyone else I’ve ghostwritten for.
He understood his audience and knew the subjects he wanted to talk about.
So he basically wanted me to focus on execution.
He would bring the expertise and ideas.
I helped package them into content people wanted to consume.
So this engagement relied more heavily on my ability to write than any other client I’ve worked with.
That being said, we still needed some structure around what we were doing.
Here’s what that looked like…
1/ We clarified the objective
The main reason he reached out to me was to save time.
But to write good content for him, I needed to understand the outcome he wanted from his content.
And this turned out to be 5 things:
A) Build authority
A major part of this engagement was maintaining and expanding my client's perception with the wider public.
He hated the thought of being seen as a content creator because there was a big stigma in the AI space at the time that most of the people creating content couldn’t build an AI system.
So he wanted to be known as the guy who builds things and creates content about it.
In other words, his content needed to reinforce his expertise in building production-ready AI systems.
B) Build an engaged audience
The next most important objective was being seen as a guide.
My client’s content was targeted at:
Practicing AI engineers
Founders
Project managers
Tech leads
Teams interested in building with AI
… and he wanted to be seen as their teacher.
Essentially, if someone got better at building AI systems because of his content, we wanted them to point back to him as someone who helped them.
C) Generate leads
When we met, my client only had a paid newsletter, but there was always an intention to build digital products in the long haul.
So we wanted to build a reliable audience of potential buyers for when a course was launched.
But my client was also interested in getting into consulting at the time, so we wanted to appeal to people looking for AI consultants.
D) Grow the wider ecosystem
LinkedIn was only one layer (and the largest) of his content ecosystem.
In 2024, he was already on Substack.
And he wanted to push some of the massive traffic he was getting on the platform towards his Substack newsletter.
E) Differentiate his content
Back then, everyone and their dog was making AI content.
But my client was adamant that he wanted his content to stand out through three levers:
Depth
Clarity
Practicality
This made his content hard to replicate, since most people at the time focused only on clarity.
2/ We created a content calendar
Like I said earlier…
This client already knew what he wanted to talk about.
The issue was everything was in his head.
Typically, I would extract this information through interviews, but over time, I realized this wasn’t going to work for this client.
Our interviews usually turned into strategy sessions, so I barely had a chance to ask questions.
Instead, I created a space in his client dashboard where he could add ideas he wanted me to turn into content.
If an idea had a deadline for publication, he would mark it in the dashboard.
This system went through several iterations before we finally landed on something:
I thought was effective for growing, nurturing, and converting his audience.
He thought would keep his content on brand.
And it looked something like this:
3 top-of-funnel posts p/w to reach a larger audience.
3 middle-of-funnel posts p/w to educate and nurture his existing audience.
1 bottom-of-funnel post p/w to push the audience toward a paid product.
This gave us structure without taking control away from him.
He still had a huge influence over what we talked about.
I just made sure every piece of content had a job and fit into the bigger picture.
