The distribution arm of the toolchain — and a bet about where the edge has moved. When anyone can build almost anything, getting it in front of the right people is the part that stays hard. The idea it's chasing: a full AI marketing team running in the background — one that learns your brand, then drafts, schedules, and posts for it.
The problem
Three things used to keep competitors out: being able to build the thing, having the money to build it, and owning data nobody else could get. All three are eroding at once. Coding has been commoditised — an AI-first builder ships in days what used to take a team months. The models themselves are a commodity: everyone rents the same frontier intelligence by the token. And data moats leak faster than they can be dug. If the ability to build is no longer scarce, building can't be the advantage.
So what's left? My bet is distribution. When two people can build the same product in a weekend, the one who wins is the one who can reliably get it in front of the right audience — over and over, without it eating all their time. Distribution is the moat that's getting harder to cross while the others get easier. Distro is that bet, made concrete.
My approach
If distribution is the differentiator, it deserves to be treated as a product to build and automate — not a chore I do by hand between builds. The end state is ambitious, and for now it's deliberately still an idea: a full AI marketing team living in the backend — agents that gather what matters about your brand, use it to draft on-message content, design it, schedule it, and post it across every channel, with no one writing each piece by hand. And it wouldn't stop at posting: the same moneyball approach my research engine takes to ideas turns inward here — watch the data for what's landing, hunt for the next opportunity, feed it back fast, and let the system get better on its own with every cycle.
The plan to get there is staged on purpose. Start with the smallest version that actually runs — a reliable, automated posting engine working on my own brands — prove that foundation, then layer the intelligence on top. Ship the narrow, real thing first; earn the right to the ambitious one.
What I built
Phase one is a multi-brand, multi-channel content-and-posting engine — the genuinely hard, valuable half of distribution, automated end to end. For any brand on any channel it drafts the post copy with one AI, paints the image with another, stamps the brand's styling onto it, holds a weekly spend cap so image costs can't run away, finds the next open slot in the posting schedule, and publishes safely — it can never accidentally post the same thing twice.
The load-bearing design choice was to build it brand- and channel-agnostic from day one. It began as the posting engine buried inside one of my own products, and the whole first phase was lifting that engine out and generalising it: the part that decides what a post is about became configuration, so the same tool can run any brand on any channel. It runs live, on a schedule, with no one in the loop. It's the dependable execution layer the AI marketing team will eventually drive — the hands, built before the brains.
Why it mattered
It's the arm of the toolchain that takes finished work to market — the counterpart to the research engine at the front. Everything upstream helps me decide what to build and build it well; Distro is what makes sure the result actually reaches people, and it does that while I'm doing something else, which is the only way distribution scales for a solo operator.
More than any single feature, it matters as a position. If the bet is right and distribution is the new edge, then owning an automated distribution engine — one that compounds every day it runs — is worth more over time than any single product it promotes.
What I learned
The valuable part turned out not to be the AI writing posts — that's the cheap, commoditised bit, exactly as the thesis predicts. The hard, durable part is the unglamorous engine around it: publishing that never double-posts, a cost cap that can't be blown, a schedule that quietly finds the next slot. Reliability is the moat, not the cleverness.
Building it phase-first kept me honest, too. It would have been easy to design the grand distribution platform on a whiteboard; instead I shipped the smallest real version, got it running on live brands, and left the sophistication for later phases to earn. The posting engine is the foundation; the AI marketing team that gathers, drafts, decides, and learns from what works is the idea it's built to carry. This is the first step of a much bigger platform — but it's a step that actually runs.