Mastermind · July 27, 2026
Distribution, Not Teachers
This week I hired two teaching assistants and roped in a few friends to help teach. The classes cover building a second brain in Obsidian and building WhatsApp AI bots, material I could teach in my sleep by now. For months I operated on the assumption that teaching capacity was the bottleneck of the education business. Find enough people who can teach this well and the thing grows.
So I solved it. Two TAs, friends lined up for guest sessions, a bench deep enough to run every cohort on the calendar.
And the growth picture didn't move at all.
On the mastermind call this morning I heard myself explain it to Xisen: the teacher is not that big of a problem. We need to build up the distribution. He was nodding before I finished the sentence, because everyone who has run a services business hits this wall eventually. I just needed to hire my way past the wrong bottleneck before I could see the real one.
What makes this sharper in 2026 than it would have been five years ago is that AI has collapsed the cost of producing the thing being sold. Any builder half a step ahead of the curve can teach Obsidian workflows or wire up a WhatsApp bot, and the curriculum scaffolding takes an afternoon to generate. Teaching supply is approaching free. What hasn't gotten cheaper, and has arguably gotten more expensive, is a reliable path to the next hundred students. Attention is the input that didn't deflate.
Xisen made a side observation on the call that stuck with me. As a kid he paid around 90 RMB a month for a VPN; today a better one costs him less, because the know-how to build them has diffused and more people are building. Prices collapse wherever knowledge spreads. Teaching AI workflows is deflating on exactly that curve, which means anyone whose plan is "be good at teaching this" is holding a depreciating asset. The appreciating asset is the audience.
Right now, I am the distribution. My posts, my referrals, my network showing up because I personally asked them to. That works, and it is also the trap: the business scales with my hours, and my hours are spoken for. This Friday I run a hackathon in Malaysia. I am also finally finishing a role application I have now dragged across two consecutive mastermind calls, which is its own embarrassment when your accountability partner keeps a list. A business that depends on my spare attention doesn't have distribution. It has a founder with a megaphone.
My north star for the business is boring on purpose: it should run reliably every month without me spending much time on it. Not a rocket ship, a machine. A machine needs three parts: product, delivery, and a channel that feeds it. Hiring confirmed the first two are solved. The channel is the missing part, and no hire fixes it by default, because a channel is an asset you accumulate, not a task you delegate.
The funny part is that other people keep trying to pay me for exactly this asset. This month alone, VCs have asked me to scout deals for them on a monthly stipend. Startup friends have asked me to be their fractional BD in the US because they list a Palo Alto address and have nobody in Palo Alto. One friend pulled me onto a customer call with a major automaker because he needed someone in the room who knew the lingo. Strip away the details and every one of these offers is the same purchase: access to rooms and people they can't reach themselves. The market keeps pricing my distribution while I keep forgetting to point it at my own product.
Xisen gave me the filter for all these side offers, and it's the line I'm keeping from this call: different options are fine, but under one theme. Said out loud, my theme is obvious. Everything I do that works, the hackathons, the communities, the scouting, is connecting people to opportunities. The education business should sit inside that theme, not next to it. Every student is someone I've connected to a capability they wanted. The channel I need to build for myself is the same one everyone else keeps renting from me.
So the assignment for the week is one line: pick the channel, own it, feed it. By next Monday's call I want to name the specific channel the next cohort of students comes from, and the answer is not allowed to be "my network." That's the founder-with-a-megaphone answer, and I've been giving it for three months.
Comments
Loading…