I Automated What Justin Welsh Does Manually — Here's What Happened

Justin Welsh makes over $5M a year from a one-person business, and his LinkedIn presence is the engine. He posts every single day, engages in the comments for 30-60 minutes, and treats LinkedIn like a full-time content channel that happens to generate revenue. He does this manually. By hand. Every day.

I respect it enormously. And I did the opposite.

I built a system that generates LinkedIn content in my voice automatically, publishes on a schedule, and frees me to spend my time on everything else. Not because I think Justin Welsh is wrong. Because I think we are optimizing for different things — and the difference reveals something important about how solopreneurs should think about content in 2026.

The Justin Welsh Model: What He Actually Does

Before I explain what I automated, let me be precise about what Welsh does. His approach is more rigorous than most people realize.

Daily posting with a content system. Welsh writes and schedules LinkedIn posts every day using a framework he has refined over years — hooks, story, insight, CTA. He has publicly shared that he batches content creation, spending dedicated hours writing posts in advance. But the posts themselves are crafted by hand. Each one.

Engagement as strategy. Welsh does not just post and leave. He responds to comments, engages with other creators, and treats LinkedIn as a relationship-building platform. He has been open about spending 30-60 minutes daily in the comments. This engagement drives the algorithm, builds audience loyalty, and generates direct business leads.

Content-to-product pipeline. His LinkedIn content feeds into a newsletter, which feeds into courses and digital products. The funnel is clean: free content builds trust, trust drives email signups, email nurtures convert to paid products. He has reportedly generated over $5M from this pipeline.

Manual quality control. Every post reflects his thinking, his experiences, his voice. The consistency of his brand is a direct result of his personal involvement in every piece of content.

It works. Obviously. The revenue proves it. But there is a cost Welsh pays that nobody talks about — and it is measured in hours, not dollars.

What I Built Instead

I did not set out to automate Justin Welsh's system specifically. I set out to solve a problem: I run three businesses, I cannot spend 90 minutes a day on LinkedIn, and I needed my content to work without me babysitting it.

Here is the system, piece by piece.

Soulin Social as the generation layer. I write one raw idea — sometimes just a sentence, sometimes a short paragraph of 60-80 words — and Soulin Social generates 35 platform-native posts, including LinkedIn-optimized content. The tool was trained on two years of my actual writing. It does not produce generic AI content. It produces content that sounds like me — the sentence rhythms, the em dashes, the dry observations, the specific details. I built this tool specifically because nothing else could do this without sounding like a robot pretending to be a human.

A scheduling and publishing pipeline. The generated posts queue into a scheduling system. I review them — this is non-negotiable, more on that later — adjust anything that feels off, and approve the schedule. Posts go out daily across LinkedIn and other platforms without me touching anything after the initial review.

An engagement monitoring bot. I built a Telegram bot — with Claude Code, because I cannot write code — that notifies me when a post gets above-average engagement. If a post resonates, I know within hours and can jump into the comments for that specific post. I do not spend 60 minutes daily in the comments. I spend 10-15 minutes on the posts that actually sparked conversation.

A performance feedback loop. Weekly, the system generates a report of which posts performed best, which themes resonated, which formats drove the most engagement. This data informs the next batch of raw ideas I feed into Soulin Social. The content gets better over time because the system learns what works — not through AI self-improvement, but through me reading the data and adjusting.

Total time investment: roughly 3 hours per week for all LinkedIn content, including review, engagement, and analysis. Welsh reportedly spends 7-10 hours per week on LinkedIn alone.

The Results: Honest Comparison

I will put real numbers here, because hand-waving about "engagement" is meaningless.

My LinkedIn (automated system):

• Posting frequency: Daily
• Average time per week: 3 hours
• Follower growth rate: Steady, organic
• Engagement rate: Consistent, occasionally spikes
• Revenue attributable to LinkedIn: Meaningful, but LinkedIn is one channel of many

Welsh's LinkedIn (manual system):

• Posting frequency: Daily
• Average time per week: 7-10+ hours (reportedly)
• Follower growth rate: Very high
• Engagement rate: Consistently high
• Revenue attributable to LinkedIn: Over $5M — LinkedIn is his primary channel

The gap is real. Welsh's manual approach produces higher engagement and significantly higher revenue from LinkedIn specifically. His personal touch — the crafted hooks, the real-time engagement, the authentic responses — creates a level of audience connection that automation cannot fully replicate. He knows this. It is why he does it by hand.

But here is what those numbers do not show: I spend 4-7 fewer hours per week on LinkedIn than Welsh does. Across a year, that is 200-350 hours. Those hours go into building product features, writing long-form content, managing KINS, and — critically — not working. Rest. Walking along the Spree. Reading. Recovering from the C-PTSD and depression that made it impossible for me to get out of bed for months before I built these systems.

Welsh chose to invest maximum time for maximum LinkedIn returns. I chose to invest minimum viable time for strong-enough LinkedIn returns, and redirect the surplus into other things. Neither choice is wrong. They are different strategies for different lives.

Where Automation Wins

There are areas where the automated approach genuinely outperforms the manual one.

Consistency without willpower. Welsh has extraordinary discipline. I do not. Before automation, my LinkedIn posting was erratic — three posts one week, nothing the next, a burst of activity followed by silence. The algorithm punishes inconsistency. My automated system posts every day whether I am motivated, sick, traveling, or having a bad mental health week. In the months after I lost motivation entirely — this was during a depressive episode in late 2023 — the system kept publishing. My audience did not notice the gap because there was no gap. That alone justified the entire project.

Multi-platform distribution without multi-platform effort. Welsh focuses primarily on LinkedIn. I publish across LinkedIn, X, Substack, Instagram, and email simultaneously because Soulin Social generates platform-native content for each. One raw idea becomes content for five platforms. Welsh's manual approach would require separate content creation for each platform — or sacrificing other channels to focus on LinkedIn. I do not have to make that choice.

Scalable content testing. Because I generate many posts from different angles on the same idea, I can test which framing resonates without spending additional time. A manual approach means each test costs creative energy. An automated approach means each test costs me 30 seconds of review time.

Removal of creative bottleneck. Some days, I do not have a LinkedIn-worthy insight. My brain is empty. In a manual system, that means no post. In my system, I can generate content from a half-formed thought — a few words, a vague direction — and refine the output. The creative threshold dropped from "fully formed insight" to "initial spark." That is a meaningful difference when you are running three businesses and your creative bandwidth is not infinite.

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

Where Manual Wins

I have to be honest about where Welsh's approach produces results I cannot match.

Depth of audience relationship. When Welsh responds personally to 50 comments, those 50 people feel seen. They become loyal. They buy his products. When I respond to 10 comments on my best-performing posts, I build 10 relationships. The math matters at scale — depth of engagement correlates with conversion rates. Welsh's audience trusts him more because they interact with him more. That trust converts at a higher rate.

Originality of each post. Every Welsh post is a unique piece of thinking. My generated posts are variations on a core idea, filtered through a trained voice model. They are good. They sound like me. But there is a qualitative difference between content that emerged from a human sitting with a blank page and content that was generated from a prompt. I can usually tell the difference. I suspect sophisticated audiences can too, even if they cannot articulate why.

Algorithm rewards for real-time engagement. LinkedIn's algorithm favors posts where the author actively engages in comments within the first hour. Welsh does this consistently. I do it selectively — only when my Telegram bot flags high-performing posts. His posts get an algorithmic boost mine do not. Over thousands of posts, that boost compounds.

Personal brand cohesion. When one person writes everything by hand, the brand voice is inherently consistent. When a system generates content from a voice model, there are occasional misses — a post that is slightly off-tone, a phrasing I would not use, a joke that does not land. I catch most of these in review. I do not catch all of them. Welsh never has this problem because he is the brand, unmediated.

The Deeper Question

Here is what this comparison really reveals: the Justin Welsh model and my automated model represent two fundamentally different philosophies about what a solopreneur's time is for.

Welsh treats content as the product. His LinkedIn presence is not a marketing channel — it is the business itself. The content creates the trust that creates the revenue. Every hour he invests in LinkedIn compounds directly into dollars. For his model, manual is not just defensible. It is optimal.

I treat content as infrastructure. My LinkedIn presence supports businesses that exist independently — KINS, Soulin Social, the membership. Content is a distribution layer, not the core product. Every hour I spend on LinkedIn is an hour I am not spending on product development, operations, or recovery. For my model, automation is not a shortcut. It is a structural necessity.

The mistake most people make is choosing a content strategy without first asking: is content my product, or is content my distribution? If it is your product — if your business model is built on audience depth and personal brand trust — study Welsh and invest the hours. If content is your distribution — if you have other products that need building and your time is the constraint — automate everything you can and redirect the surplus.

I chose automation because I physically could not sustain the manual approach. Not because I lacked ambition — because I have C-PTSD and ADHD and three businesses that need different kinds of attention, and 90 minutes of daily LinkedIn engagement would have broken me. The automated system did not just save time. It saved my health. That is a variable most productivity content conveniently ignores.

How to Decide for Yourself

If you are trying to figure out which approach is right for you, here is the honest framework:

Go manual if: Content is your primary product. You have 10+ hours per week to dedicate to one platform. You enjoy the engagement process. Your business model depends on personal trust and audience relationship depth. You are building a personal brand as the core business asset.

Go automated if: Content supports a product-based business. You have fewer than 5 hours per week for content. You need to be on multiple platforms. You value time flexibility over maximum engagement. You have other priorities — health, family, other projects — that matter as much as content metrics.

Go hybrid if: You want the best of both. Write your highest-stakes posts manually. Automate the rest. Engage personally on your top-performing posts. Let the system handle distribution. This is what I actually do now — about 20% of my posts are written entirely by hand for strategic moments, and 80% come through the automated system.

The tools to automate are available now. Soulin Social does the generation. Claude Code builds the bots. Telegram handles the notifications. The total cost is under $150/month. The time savings are 200+ hours per year.

But the choice is not about tools. It is about what you want your days to look like. Welsh chose mastery of a single channel. I chose freedom across all of them. Both are making money. Both are building alone. Both are right — for the person making the choice.

What matters is that you choose deliberately, knowing the trade-offs, instead of defaulting into one model because some article told you it was better.

Now you know both. Choose the one that fits the life you are actually building, not the one that looks best in a case study.