Brunson tells you to do something that feels wrong.
Build the marketing materials for your class before you build the class.1
Most people design the product first, then figure out how to sell it.
Brunson calls that "a formula for disaster" and reverses it — write the marketing, then build the thing to match.2
It sounds backwards because it is, deliberately.
The instinct is that you can't market something that doesn't exist yet. Brunson's counter is that marketing-first is exactly how you avoid building something nobody wants.
The logic is about which mistake is cheaper to fix.
If you build the product first and then try to market it, you've committed months and money before discovering whether anyone wants it.
If the marketing won't come together — if you can't write a headline that excites anyone — you've already sunk the cost, and you find out too late.
If you build the marketing first and it won't come together, you've lost a few hours and learned the idea doesn't sell before building anything.
The marketing is a cheap test of the expensive product.
So marketing-first isn't really about marketing. It's about validation sequencing — putting the cheap test before the expensive commitment, so you fail fast and cheap instead of slow and dear.
There's a second function: the marketing becomes the specification for the product.
When you write the headline, the promise, the hooks first, you've committed to what the product must deliver. The product then gets built to fulfill the marketing's promise, rather than the marketing getting stretched to describe whatever you happened to build.
This keeps the product honest to a customer-facing promise instead of drifting toward what's easy or interesting for the creator to make. The marketing says "this delivers X without Y"; now the product has to actually do that.
Brunson's sequence bakes this in: Who/What statement, then headline, then curiosity hook, then Ask Campaign, then build the class to answer the questions. Every product-defining decision is made from the customer's side first.
The clearest instance is the Ask Campaign sitting before the curriculum.
You build the one-question page (marketing/research), collect responses, and only then design the modules from the recurring questions. The product's structure is literally dictated by the pre-product research.3
Notice what this prevents. Without it, the expert builds the curriculum they think is right, then discovers (maybe) that the market wanted something else. With it, the curriculum can't be wrong about what the market wants, because it's built from what the market asked.
This is marketing-first in its strongest form — not just writing copy early, but letting the pre-product marketing determine the product. The class exists to answer questions the marketing already collected. The order isn't just efficient; it makes the product structurally responsive to demand.
The principle has a failure mode: marketing-first can optimize for what's easy to sell over what's valuable to deliver.
If you build the marketing first, you naturally write the most exciting, most sellable promise. Then you build a product to match that promise — which is good if the exciting promise is also a good product, and bad if the most sellable angle isn't the most valuable one.
The marketing-first sequence pulls toward sellability, and sellability and value aren't identical. A product built to fulfill the sexiest possible headline may be worse than one built to genuinely help, if the sexiest headline overpromises.
Brunson's system has no correction for this because it's optimizing for conversion. The honest operator uses marketing-first to validate demand and avoid building unwanted things — while making sure the promise they lead with is one the product can actually keep. Marketing-first that leads with a promise you can't fulfill just fails you faster in the other direction: a sold product that disappoints.
The resistance to this sequence is worth understanding, because it's almost universal and almost always misguided.
Building the product first feels responsible. You're making the real thing, doing the hard work, being substantive rather than "just marketing." Writing marketing for something that doesn't exist feels like putting the cart before the horse, even dishonest.
But that feeling mistakes effort for progress. Months spent building an unwanted product is not responsible; it's expensive failure that felt like diligence the whole way down. The marketing that "feels premature" is actually the cheapest possible test of whether the months are worth spending.
The discomfort is the sunk-cost instinct running in reverse — you want to build first because building feels like commitment, and commitment feels serious. Marketing-first asks you to earn the right to commit by proving demand first, which feels like skipping ahead but is actually refusing to gamble months on an unvalidated guess. The wrongness you feel is the correct sequence fighting your instinct to mistake labor for validation.
You're about to spend months building a product.
Don't. Write the marketing first — the Who/What statement, the headline, the hook. Spend a few hours, not a few months.
Watch what happens as you write. If the headline excites you and would plausibly excite the market, that's signal the idea has legs. If you can't make it compelling no matter how you phrase it, that's the idea telling you it won't sell — and you just learned it for the cost of an afternoon.
Then run the Ask Campaign before building the curriculum, so the product's structure comes from real demand rather than your assumptions.
Build the product to fulfill the marketing's promise — but check that the promise is one you can actually keep. Marketing-first protects you from building unwanted things; it doesn't protect you from overpromising. Lead with the most compelling promise your product can honestly deliver, not the most compelling promise you can write.
Marketing-first as validation: you write the marketing to test whether the idea sells before building, and you lead with a promise the product can genuinely keep. The cheap test precedes the expensive commitment, and the promise stays honest.
Marketing-first as overpromise: you write the sexiest possible marketing, then build a product straining to match a promise it can't quite keep. You've optimized for sellability over value, and the product disappoints on delivery.
The difference is whether the pre-built marketing is testing demand honestly or writing checks the product can't cash. Both are "marketing-first"; only one is a validation tool.
Asserted from experience, no data.4 The underlying principle (validate demand cheaply before building) is sound and echoes lean-startup and pre-sell methodologies well outside this book, though Brunson cites none.
Tension: marketing-first optimizes for what sells, which pulls toward the most exciting promise, which isn't always the most valuable product. The sequence protects against building unwanted things and can push toward overpromising, and Brunson only names the first benefit.
Second tension: "build marketing first" assumes you can write honest marketing for something that doesn't exist. For a genuinely novel offering, you may not know what you can deliver until you build it, making the pre-built promise a guess you're now committed to.
Open question: where the most sellable promise and the most valuable product diverge, does marketing-first systematically bias operators toward the former — and is there a version of the sequence that validates demand without pulling toward overpromise?
Convergence with the Ask Campaign is total — the Ask Campaign is marketing-first made concrete, with the pre-product research determining the product. It also connects to the Who/What and headline pages, which are the marketing you build first.
The tension with the vault's craft corpus is real. Craft-first thinking (make the work good, then find its audience) is the opposite sequence, and it's right for art in a way it may not be right for information products. Brunson's marketing-first suits a product whose value is solving a stated customer problem; it suits worse a work whose value the audience couldn't have specified in advance. Both sequences are correct for their object.
To The Ask Campaign. The Ask Campaign is the strongest instance of marketing-first — research and marketing conducted before the product, then determining it. This page states the principle; the Ask Campaign is the flagship implementation.
Held together: marketing-first isn't just "write copy early," it's "let the market define the product before you build it" — and the Ask Campaign is how you do that concretely. The principle without the method is just a nice idea about sequencing; the method without the principle is a research tactic without a rationale. Together they form the complete argument: build the cheap customer-facing thing first, let it dictate the expensive thing, and you can't build something unwanted.
To Why Improvement Offers Fail. Marketing-first surfaces whether an idea sells before you build it — and one thing it surfaces is whether you've accidentally designed an improvement offer. If you can't write an exciting headline, you may be looking at an improvement offer's fatal dullness.
The insight neither reaches alone: the marketing-first test is also a new-opportunity-vs-improvement-offer test — an idea that won't yield a compelling headline is often an improvement offer in disguise, and the failed marketing is the early warning. The improvement-offer page explains why certain offers won't excite; marketing-first is when you find out — before building, cheaply, at the headline stage. The dead headline and the improvement offer are frequently the same diagnosis arriving early.
Sharpest implication. Marketing-first isn't about marketing — it's validation sequencing: the marketing is a cheap test of the expensive product, so you fail fast and cheap instead of slow and dear, and in its strongest form (the Ask Campaign) the pre-product research actually determines the product's structure. But the sequence has a shadow: it pulls toward the most sellable promise, which isn't always the most valuable product, and it protects against building unwanted things without protecting against overpromising. Used honestly it prevents the wrong product; used carelessly it commits you early to a promise you can't keep.
Generative questions.
Where the most sellable promise and the most valuable product diverge, does marketing-first systematically bias toward sellability — and can the sequence be run so it validates demand without pulling toward overpromise?
If a failed headline is often an improvement offer in disguise, is the deepest function of marketing-first not efficiency but positioning diagnosis — a way to catch, before building, that you've designed something structurally hard to sell?