Why I'm burning down my $1M course business to restart my agency: chapter one.
This is the story of how I built the business of my dreams, fell for the promise of the course creator model and made a million dollars, then burned it all down to build my dream business back up again from scratch.
From 2003 - 2020, I built up a whole freelance business. From selling my first email copywriting package for $20, all the way up to landing big projects for me and my team to work together. I built it all based on relationships, word of mouth, networking, and product-led growth. We had the lowest churn and highest LTV of any agency I knew.
And then at the tail end of 2020, I made what I thought would be the best decision of my life. I bought into the fairytale I'd been seeing for all those years, and I decided to remove myself from client work and become an online course creator.
For four glorious years, I did exactly that.
I carved out a nice little home for myself, teaching the methods I had used to get me clients all those years. Specifically, teaching them to neurodivergent people. People like me, who couldn't be consistent but really, really needed consistency from their businesses. I did a little over $1M in revenue in that time, made a ton of friends, built a loyal audience and following, and thought I had made it.
Nothing could have prepared me for what would happen next.
AI and LLMs replaced courses for a lot of people. They could never replace community or connection, so even though I lost a lot of the course market? It was a minor setback that required some tweaking of my offers. It would have been fine.
Except my key target market from 2021-2025 was:
- Mid-market freelancers and consultants
- Primarily writers, designers, admin and ops contractors, and marketers
- B2B services targeting sub-$1M companies and solo operators
- And neurodivergent, specifically.
Take a wild guess at the very first major sector to experience the work displacement and replacement of the AI era. Allow me to list off some rather depressing stats for you, in case you fall into this group and you're still blaming yourself for what happened:
- Freelance writing jobs dropped by 30%, software/translation/admin work dropped by 20%, and design work fell by 17%, while Upwork and Fiverr spend from corporate plummeted nearly 80% since 2022. That's not a drop, that's a cliff.
- Over half of small businesses that had hired freelancers in 2022 had completely stopped outsourcing those roles by late 2025.
- And all of this specifically harmed neurodivergent freelancers and consultants more, because it tripled the time and administrative cost of generating leads. So you spent 3x as much time and energy just getting clients, before even doing a single hour of client work. When executive dysfunction and RSD make client attraction hard, something that triples that load and makes it more admin heavy? Basically a death blow.
At this point, I don't go a week without seeing 3-5 of my former colleagues, subscribers, and clients posting about the new job they've landed. And these are brilliant, talented people who spent years with a steady stable of clients.
I'm proud of them for doing what they have to do to survive this pivot, while also feeling genuinely bad that they have to. But also?
I wasn't exactly blindsided by this, but I held on too long.
This market, this business, the thing I had built serving the people I had grown to admire and appreciate? It was my darling. And the hardest thing in the world is for a founder to kill their darlings.
So, for 2025, and so far in 2026, I held on.
I ran new, more creative promos. I revamped my offer suite into the most effective set of tools I've ever built, including a suite of micro-tools for service-based solo operators, a community for new micro-SaaS founders, and even a micro-SaaS of my own. And then...
... I marketed them all to the same audience that was struggling and suffering and going back to corporate at a constant pace, because I didn't want to admit that no amount of retooling an offer for a market that quite simply can't afford it (even if they really like me and really want it) will magically fix things.
Have I still done six figures per year through 2025 and 2026? Yeah. Rolling 12 months back right now we're at about $180k, and since July 2024 we're close to $400k. In this market, for a solo operator targeting a truly struggling clientele?
There is no doubt that my marketing is good, my conversion is solid, my copy works, my messaging and positioning are on point, and my offer is in demand.
But.
Is it stable? Is it predictable? Is it growing?
Does it feel good?
No, no, no, and nooooooo.
I miss agency life, to be honest.
I miss building a team, hiring super competent professionals, leading a visionary strategy that actually gets results. I miss sticking around with a business long enough to see those results happen.
I miss being someone's secret weapon.
I miss the stability of a 6-month project where I could really, truly give it my focus instead of fragmented attention across dozens or hundreds of clients.
I'm not dropping any of my current projects, of course. But I am going to bring the agency back, and to do it I'm going to do something very specific.
I'm going all-in on my agency, and using the exposure from growing it to market my other products downstream - and documenting the entire thing here for you to see.
The agency: Tactile Design Co - human-led agentic marketing for micro-SaaS
The tools: Alex (content distribution and lead gen), Solo School (micro-tools for neurodivergent service providers), and the Craft Software Collective (build kits and community for pre-revenue micro-SaaS.)
They all serve the same market at different stages.
Knowledge workers and experts start off offering services, get an idea for a micro-SaaS and start building, then need some automation and support, and then need an expert to take over.
So my content and my tactics? Feed all of them at once.
But in 2026, you can't slap up a fucking blog post and watch the world beat a path to your door. Algorithms don't care about you, email is getting read by chatbots instead of humans, search has been replaced by a chatbot, ads are more expensive than ever and have their own content hamster wheel of doom. The playbooks from 2018, 2020, 2022, and even 2025 don't work anymore.
Here's what I'm focusing on, and will be documenting here:
- AEO - if word of mouth and referrals are the most powerful marketing strategy on earth, recommendations from a chatbot that the user trusts with literally everything? Just as powerful, if not more so. And if you're not showing up in those results, you're not showing up for buyers anymore. Period.
- Zero-click content - algorithms will bury just about anything that tries to take people off platform, because if a user leaves, they can't be shown more ads. Zero-click content that builds trust and reach and gives value without leaving the platform is the only type of content worth creating in 2026.
- On-platform lead capture - to that same end, catching leads on platform (whichever platform that is) is going to out-convert multi-stage funnels. There are ways to do this, most of which are totally above board and some of which I will only ever use on my own accounts for TOS reasons.
- DM funnels - if you want on-platform lead capture that turns into email marketing? DM funnels still work, especially if you can automate the parts that no one expects to be human and then bring in a human at the right moment.
- Channel partnerships - a rising tide floats all the boats. Affiliate partnerships, podcast guesting, all forms of earned media, joint ventures, collaborations... Very few forms of marketing are more powerful than an implied endorsement to a captive audience.
- Product-led growth - lowering churn, increasing LTV, adding virality and dark social moments to tools, apps, communities, lead magnets and funnels. The best marketing for your offer is it's fans, the second best marketing is the offer itself.
With every single one of these, I'm going to give you my exact moves. The setup, the real numbers, the wins and losses, the agents and workflows.
If you want the tools that run it all? Well, I've built a whole bunch of those you can buy, if you want the help. But if you just want to build them yourself?
Make sure you're following along, because I'm not leaving anything out.