How I Built an Automated Content Flywheel That Runs While I Sleep

In January 2025, I published 8 pieces of content across all platforms. Total. For the entire month. I was writing each post from scratch — a LinkedIn post here, a tweet there, a newsletter when I had the energy. I was averaging maybe 45 minutes per post, which meant content alone was eating 6 hours of my week, and the output was embarrassingly thin.

By March 2025, I was publishing 35+ pieces of content per week. Same hours. Same me. No ghostwriter. No content team. No VA in the Philippines writing tweets at 3am.

The difference was not working harder. The difference was building an automated content flywheel — a system where one single idea enters at the top and comes out the bottom as a full week of platform-native content, formatted, scheduled, and ready to post.

This is not a concept. This is a machine I built, broke, rebuilt, and now run every single day. Here is exactly how it works.

Why Most Content Systems Fail Solopreneurs

I need to say this before I explain the system, because I wasted months learning it: most content advice is written for people with teams.

"Batch your content." Great — when you have a content manager to edit, a designer to format, and a scheduler to publish. When you are one person, batching just means doing all the exhausting work on one day instead of spreading it across five. The exhaustion is the same.

"Repurpose everything." Also great in theory. In practice, "repurposing" a blog post into a tweet takes genuine creative effort. You cannot just copy-paste the first paragraph. Each platform has different norms, different formats, different audiences. Manual repurposing is almost as slow as creating from scratch.

The content flywheel only works when the transformation from one idea to many formats is genuinely automated — not "automated" meaning you still do it yourself but faster, but automated meaning the system does it while you do something else.

That is what I built. And it took me three attempts.

Attempt One: The Manual Spreadsheet (Failed)

My first content system was a Google Sheet with columns for each platform. I would write a core idea, then manually adapt it for LinkedIn, X, Instagram, Substack, and email. I color-coded everything. It was beautiful. It was also completely unsustainable.

The problem was not the spreadsheet. The problem was me. I am a person with ADHD who dropped out of university and healed from 15 years of C-PTSD. My executive function has limits. A system that requires me to make 35 micro-decisions per content cycle is a system designed for someone with a different brain.

I lasted six weeks before the spreadsheet became a monument to guilt — rows of empty cells staring at me every morning.

Attempt Two: The Tools Frankenstack (Failed Better)

Next I tried chaining tools together. Write in Notion. Reformat with ChatGPT. Schedule with Buffer. Track with Airtable. Connect with Zapier.

Five tools. Five monthly subscriptions. Five things to maintain, update, and debug when they inevitably broke. The Zapier automations failed silently about once a week — I would discover three days later that nothing had posted to LinkedIn because a webhook timed out. Buffer's scheduling interface worked fine until it randomly logged me out and ate a week of scheduled posts.

This version was better than the spreadsheet. I was publishing more. But the maintenance cost was brutal. I was spending almost as much time fixing the pipeline as I was creating content.

Attempt Three: The Flywheel That Actually Works

The version that works — the one running right now — has exactly three components.

Component 1: The Idea Capture

Every morning at 8am Berlin time, a Telegram bot sends me a curated list of trending topics in my niche. This is a Node.js process I built with Claude Code, running on PM2. It monitors X, LinkedIn trending topics, Google Trends, and a few niche newsletters. It scores each topic for relevance to my audience and sends me the top five.

I do not always use these topics. Sometimes I wake up with something on my mind — a conversation from the day before, a pattern I noticed, a mistake I made. But on the mornings when I have nothing, the bot ensures I am never staring at a blank page.

Time spent: 5 minutes scanning the list, choosing a direction.

Component 2: The Raw Thought

This is the only part that requires my actual brain. I write a raw thought — anywhere from 50 to 150 words. Unpolished. No formatting. Just the core idea.

Here is a real example from last week:

"I spent $480/month on 14 SaaS tools and got maybe $60 of value. The problem wasn't the tools. It was that I was using tools to feel productive instead of being productive. Every tool you add is a tool you have to maintain. The best tech stack shrinks over time."

That is 52 words. It is sloppy, it is raw, and it is the seed of everything that follows.

Time spent: 10-15 minutes.

Component 3: The Multiplication Engine

This is where the flywheel spins. I feed that raw thought into Soulin Social — the tool I built specifically to solve this problem — and it generates platform-ready content in my voice.

From that 52-word seed, I get:

• 3 LinkedIn posts (different angles, different hooks, different formats — one story-driven, one tactical, one contrarian)
• 5 tweets (standalone insights, each under 280 characters)
• 2 tweet threads (longer breakdowns, 4-7 tweets each)
• 3 Instagram captions (with relevant hashtag sets)
• 1 Substack intro paragraph
• 2 email newsletter snippets
• A blog post outline with key arguments expanded
• Platform-specific variations I did not even plan for

The key is voice fidelity. I trained this on two years of my actual writing. It knows I use em dashes. It knows I do not use exclamation points on serious sentences. It knows my sentence rhythm — short, short, long, short. It captures the warmth-with-a-blade-underneath energy that is just how I write.

I review everything. I adjust maybe 20% of the output — tightening a hook here, adding a specific detail there. But 80% goes out as-is.

Time spent: 20 minutes reviewing and adjusting.

The Math

Total daily time investment: 35-40 minutes.
Total weekly output: 35+ pieces of platform-native content.
Total monthly content volume: 140+ posts.

Before the flywheel: 6 hours/week for 8 pieces/month.
After the flywheel: 3 hours/week for 140+ pieces/month.

That is a 17x increase in output with a 50% decrease in time. Those numbers are real. I track them.

The Parts Nobody Talks About

Distribution Is Not Automatic (Yet)

I still manually post most content. I know — for someone writing about automation, this feels hypocritical. But here is why: each platform's algorithm rewards different posting times, different formats, and different engagement patterns. Automated scheduling tools post at predetermined times regardless of what is happening on the platform that day.

I am building automated distribution into Soulin Social. It is not ready yet. When it is, the flywheel will be fully autonomous — idea to published to tracked, zero manual steps. For now, the multiplication is automated but the distribution is semi-manual. Honest answer for an honest guide.

Quality Control Matters More Than Volume

In my first month running the flywheel, I published everything it generated. Volume was intoxicating. More posts. More reach. More visibility.

Get essays like this — plus behind-the-scenes builds — in your inbox every week. Subscribe free →

Then I looked at the engagement data. The low-quality posts — the ones I should have edited or killed — were actively hurting my average. Platforms punish accounts that post mediocre content. My reach actually decreased during the high-volume phase because the algorithm started showing my posts to fewer people.

Now I kill about 15% of what the flywheel generates. Some ideas do not translate well to certain platforms. Some posts are technically correct but lack energy. Deletion is part of the system.

The Long-Form Feeds the Short-Form

The flywheel works best when it has long-form content to draw from. My weekly Substack essay — typically 1,500-2,500 words — becomes the richest source material. One essay generates enough short-form content for an entire week.

This creates a virtuous cycle: the short-form content drives traffic to the long-form, which drives subscriptions to the newsletter, which gives me more raw material for the flywheel. Content feeding content feeding content. That is the flywheel in "flywheel."

What This Actually Costs

The flywheel itself costs very little in tools:

The real cost is the upfront investment. Building Soulin Social took months. Training it on my voice took weeks of iteration. Building the content scout bot took a weekend. The system did not exist on day one — it was assembled piece by piece, each component built after I understood the manual version well enough to know what "good" looked like.

If you do not want to build your own tool, you can use Soulin Social directly. That was the whole point of making it a product — so other solopreneurs could skip the two years of duct-tape solutions I went through.

The Mistakes That Taught Me Everything

Mistake 1: Automating before understanding. I built my first content system before I understood what good content looked like. The result was efficient distribution of mediocre work. The flywheel amplifies whatever you feed it. Feed it mediocre ideas, get mediocre content at scale. I spent three months publishing garbage efficiently before I realized the input quality was the bottleneck, not the distribution.

Mistake 2: Ignoring platform differences. A LinkedIn post is not a tweet with more characters. An Instagram caption is not a LinkedIn post with hashtags. Each platform has different norms, and violating them tanks your reach. The flywheel generates platform-native content specifically because I learned this lesson the expensive way — by watching identical cross-posted content perform 80% worse than native content.

Mistake 3: Chasing virality. I once had a tweet go semi-viral — 200K impressions. It brought me exactly zero customers. The content that builds a business is not the content that goes viral. It is the content that attracts the right 500 people and makes them think, "This person understands my problem." The flywheel optimizes for resonance, not reach.

How to Build Your Own (Even If You Are Not Technical)

You do not need to build a custom tool. Here is the minimum viable content flywheel:

1. Set a recurring alarm for idea capture. Every morning, before you open email, write one raw thought. 50-100 words. The idea, unpolished.

2. Use an AI tool to multiply. Claude, ChatGPT, whatever you prefer. Prompt it with your raw thought and ask for platform-specific variations. Give it examples of your past posts so it matches your voice.

3. Review ruthlessly. Kill anything that does not clear your quality bar. Volume without quality is noise.

4. Post natively. Do not cross-post identical content. Each platform gets its own version.

5. Track what works. After a month, you will see which types of content perform on which platforms. Feed that data back into the system.

This manual version takes about an hour a day instead of my 35 minutes. But it works. And it is infinitely better than the "write every post from scratch" approach that burns out every solopreneur I know.

When you are ready to go further — full automation, voice training, one-click multiplication — that is what Soulin Social is for. And the complete workflow, including how it connects to the rest of my operating system, is inside the membership.

But start with the alarm and the raw thought. That is where the flywheel begins. Not with a tool. With a decision to show up every morning and think one thought worth sharing.

The rest is machinery. And machinery, I have learned, is the easy part.