A car company decides to enter a new country. The usual move is to ship cars, open dealerships, and pray people show up. Koe's move is stranger: first you publish the best driving guide in that country, build a following of people who trust your taste in roads, and only then roll the cars onto a market that already knows your name. The guide is cheap to make and prints money. The cars can lose money for a year, because the guide is doing the work of finding the buyers.
That's the wedge. Koe describes building a software company by taking it to market through an education product first — skipping what he calls the sink-or-swim phase of a normal software launch.1 He says he plans to take the company to 5–10 million in its first year not by selling the software but by selling an education product around the software's subject, running at 95% profit margins, with the company's social media built like a personal brand fronted by an avatar.2 The software gets injected slowly, even if it isn't profitable in year one.3
Underneath the specifics is a sequencing rule. When the real product is expensive, slow, or fragile to launch cold — software being the example — you don't lead with it. You lead with a cheap, high-margin information product that attracts and educates the exact people who'll later want the hard product.4
The education product does three jobs at once. It funds the operation while the hard product is still unprofitable.5 It builds the distribution — an owned audience — that the hard product would otherwise have to buy.6 And it pre-educates buyers into wanting the category, so the hard product lands on warm ground instead of cold.7 Koe is explicit that you can run this without any software at all; the software is optional, the education layer is the engine.8
Here's the deconstruction. Software is sticky in the bad way — churn, developer costs, the grind of getting people to use it daily.9 An education product has none of that friction: near-pure margin, no servers to keep an app alive, and it sells to a mind rather than a workflow.
So you use the frictionless thing as a battering ram for the friction-heavy thing. Imagine a band that can't fill a stadium yet. They don't book the stadium. They release a free mixtape, sell out small rooms, build a list of people who'd follow them anywhere — and then the stadium is a formality. The mixtape de-risked the stadium. Koe's education product is the mixtape; the software is the stadium.
The avatar-as-face detail matters more than it looks.10 Koe says most businesses have no idea how to build a real social audience, so they rent one through creator sponsorships.11 By fronting the company with a personality and building it like a personal brand, you grow your own distribution instead of renting it forever. The wedge isn't just "sell a course first" — it's "sell a course first so you own the channel the software will ride in on."
Several vault pages argue that owning your distribution beats renting it; this page supplies the order of operations that turns that thesis into a launch plan. It pairs directly with Own Distribution vs Manual/Bot/Borrowed — that page taxonomizes the channels, this one shows how an education product manufactures the "owned" column from scratch before a harder product needs it.
It also operationalizes Every Business Needs an Education Layer by making the education layer not just a feature but the first thing you ship. And it gives a concrete launch shape to Customer Creation Through Education: the wedge is how you create customers for a product that doesn't have a market yet. The sibling Play the Game to Be Free of the Game supplies the deeper posture — you run the unprofitable software patiently because the education layer buys you the freedom to play long.
The cleanest case is the one Koe is in the middle of as he speaks. He won't name the software — it's still in legal review — but he narrates the entire wedge.12 The structure he lays out: education product as the main revenue driver in year one, company social media built like a personal brand with an avatar face, that audience plus a "plethora of creator audiences" funneled into a cohort-based course about the software's subject, and only then the software injected slowly as "the one true software" of its niche.13
What's worth dissecting is the order and the patience baked in. The course around the subject of the software runs before the software is the product. The cohort isn't a customer-acquisition afterthought; it's the on-ramp that makes the software's eventual arrival feel inevitable to an already-educated room. And the explicit permission to be unprofitable on the software in year one only works because the education product is carrying the P&L.14 Strip the education wedge out and this is just another underfunded software launch fighting for cold attention. With it, the software arrives to a built audience and a funded runway.
It's 9:00 on a Monday. The software exists only as a Figma file and a half-built backend. You don't open either. You open a doc titled Course — Module 1 and start writing the thing your future users will need to understand before they'd ever pay for the tool. You record a short video for the company account — not a product demo, there's no product yet, just you explaining the problem the way a friend would.
By the afternoon you've got two paying cohort members and a Notion page of objections they raised on the call. Those objections are your product roadmap. The codebase can wait another month; the audience can't. When the software finally ships, it ships to a list of people who already trust the voice that's been teaching them — and the launch feels less like a cold open and more like a sequel.
You'll know the wedge has snapped when:
The biggest flag is structural: this is a forward-looking plan, not a result. 🚩 As Koe records, the software is unnamed and unreleased, in legal review.15 The "5–10 million in year one" is a stated goal, not an outcome — treat the whole wedge as a hypothesis from an experienced operator, not a proven playbook.16 [POPULAR SOURCE] 🚩 MOTIVATED REASONING — Koe sells education products; an argument that education products are the ideal go-to-market is also an argument for what he already does.
A genuine tension: Koe claims the software can be unprofitable in year one and that's fine. That's only fine if the education product's margins are as fat and as durable as he says — and the "95% profit margin" figure is asserted, not sourced.17 If the education revenue softens, the unprofitable software stops being patient strategy and becomes a hole. The plan has a single point of failure it doesn't flag: the wedge funds the operation, so if the wedge wobbles, everything downstream wobbles.
Open question: does the wedge generalize past Koe's own niche (creators), where his personal audience already primes the funnel — or does it require a pre-existing audience most operators don't have?
Koe and Jack Moses both prize patient, owned-channel building, and the convergence is clean with Play the Game to Be Free of the Game. Moses's posture — endure the short-term constraint to buy long-term sovereignty — is exactly what lets Koe say "the software can lose money in year one." The education product is the game played patiently so the software can be free of the usual launch desperation. Where they'd diverge: Moses tends to start from the solo operator with little capital, while Koe's wedge quietly assumes enough runway to float an unprofitable software product on the side. That's a richer-operator move dressed in solo-creator clothes, and it's worth naming the gap rather than pretending the advice is equally available to everyone.
There's also a friendly clash with the market-research orthodoxy that Koe spends the rest of the source attacking. Standard advice says find a starving market and build for it. Koe's wedge says create the market by educating people into wanting the category, then sell them the product the education primed them for. He's closer here to the create-don't-find-a-niche instinct than to classic demand-led product strategy — and the cohort course is the create-the-market mechanism in action.
Plain version: instead of launching your real product cold, you first sell a cheap course that teaches people why they'd want it — and that move connects to how positioning experts say you should create a market rather than hunt for one, and how patient operators trade short-term profit for long-term control.
The first handshake is with Perell — Create, Don't Find a Niche. Perell's argument is that the durable position is one you invent rather than one you discover already crowded. The education wedge is the temporal version of that idea: it doesn't just create a niche in concept-space, it creates a cohort in time — a body of pre-educated buyers who didn't want the category until the course taught them to. Put the two together and a sharper claim falls out: creating a niche isn't only a positioning act, it's a sequencing act. You create the niche by educating an audience into it before the product that serves it exists, so that demand and supply are manufactured in the right order. Perell tells you to create the position; the wedge tells you the education product is the tool that builds the audience that makes the position real. Neither idea alone gives you the timing rule — create the market with a course, then drop the product into the demand you built.
The second handshake reaches to Play the Game to Be Free of the Game in the Moses cluster. That page is about enduring a constraint now to win freedom later. The wedge is a precise instance: you accept an unprofitable software product (the constraint) because the education product (the game, played well) funds and de-risks it toward a position no cold launch could reach (the freedom). The insight that emerges from holding them together is about which thing you let lose money. Most operators protect the flagship and starve the marketing. The wedge inverts it — you let the flagship lose money and you make the marketing layer itself the profit center. The education product isn't a cost of customer acquisition; it's the business, temporarily, so the flagship can be patient. That inversion — marketing-as-profit-center funding flagship-as-loss-leader — is something neither the patience page nor the launch page states on its own, and it's the actual structural trick.
The Sharpest Implication. The education product isn't a marketing expense — it's the part of the business that's allowed to make money first, so the "real" product can afford to be slow and bad in year one. If that's right, then the question for any hard-to-launch product flips from "how do I sell this?" to "what's the cheap, high-margin thing I can sell that pre-educates the buyers and funds the runway?" The wedge reframes the whole launch as a two-product sequence where the easy product carries the hard one.
Generative Questions.