After you run the Ask Campaign and collect about 100 responses, Brunson says something useful happens: the curriculum is already written.1
Out of 100 answers, 8–10 core questions recur.2 Those recurring questions become the module titles of your class. You don't design the curriculum; you extract it from what people actually asked.
The move converts market research directly into product structure.
The questions people typed are the syllabus.
Brunson's own example is the potato gun product, and it makes the extraction concrete.3
The Ask Campaign for "how to make a potato gun" produced questions like these:
What type of pipe do you use? Do PSI ratings matter? Which igniter — barbecue or lantern? How long to cut the pipes? What propellants? Am I going to die if I make it wrong?
He grouped the recurring ones into modules: How to Decide Which Gun, How to Pick the Right Pipes, How to Find Accessories, How to Cut the Right Ratio, How to Assemble, How to Increase Distance, How to Stay Safe.4
Notice the last one — "Am I going to die?" became a whole safety module. The audience's fear, in their words, became a curriculum unit.
That's the method working: the questions carry not just topics but the emotional weight the audience attaches to them.
The extracted curriculum has an advantage the designed one can't match: it can't be wrong about what the audience wants to know.5
A designed curriculum reflects what the expert thinks matters — which, per the Ask Campaign's whole premise, is systematically off from what the audience actually wants.
The expert would probably lead with ballistics theory; the audience wants to know if they'll die.
An extracted curriculum reflects what the audience asked, in their order of concern. The 8–10 recurring questions are, by definition, the things enough people care about to type. Building the class around them guarantees relevance.
There's a second benefit: the questions are already in the audience's language, so the module titles sound like the audience rather than like a textbook.
"Am I going to die if I make the gun wrong?" is a better module title than "Safety Considerations in Pneumatic Device Construction," and it came free.
The number of responses matters, and 100 is doing specific work.
With 20 responses, you can't tell a recurring question from a one-off. Two people asking the same thing might be coincidence. The signal — which questions most people care about — hasn't separated from the noise yet.
By 100, the recurring questions have surfaced clearly. The 8–10 that keep appearing have appeared enough times that you can trust they're real patterns, not accidents. The volume is what makes the frequency meaningful.
There's a diminishing-returns logic too. The first 100 responses reveal most of the recurring questions; the next 100 mostly repeat them. So 100 is roughly where you've captured the pattern without wasting effort collecting redundant data. It's the point where the curriculum has stabilized.
Extraction hands you not just the modules but their sequence, which is easy to miss.
The frequency and phrasing of questions reveal what the audience wants to know first. In the potato gun case, foundational questions (which gun, which pipes) naturally precede advanced ones (increasing distance), because that's the order concern actually arises.
You don't have to design the learning sequence either — the audience's questions imply it. What they ask early and anxiously comes first; what they ask as refinement comes later. The safety question ("am I going to die") sits where the audience's fear puts it, which is often earlier than an expert would place it.
So extraction gives you structure and order. The 100 responses are a fully-sequenced outline, not just a topic list, and the sequence is the audience's own priority order rather than the expert's logical one.
The method's blind spot deserves its own treatment, because it's where extraction can actively harm.
An extracted curriculum covers what people knew to ask. It cannot cover what they didn't know to ask — the non-obvious factor, the hidden prerequisite, the danger the audience is unaware of. The potato gun buyers asked about propellants; if there were a subtle failure mode nobody suspected, no one would have asked, and pure extraction would omit it.
This is the same blindness the audience has, faithfully reproduced. The curriculum inherits the audience's ignorance about its own ignorance, which is precisely where the most important teaching sometimes lives.
The fix isn't to abandon extraction — it's to extract the curriculum, then have the expert audit it for the critical thing no one asked about, and add that deliberately. Extraction sets the structure; expertise patches the unknown-unknowns. Skipping the audit means teaching a class that answers every question the audience had and misses the one they should have had.
You have ~100 Ask Campaign responses and need to build the class.
Read all the answers and tally the recurring questions. You're looking for the 8–10 that keep coming up — those are your modules, chosen by frequency, not by your judgment of importance.
Group related questions under each recurring theme. Sub-questions under a module become the lesson points within it, again in the audience's words.
Resist the urge to add your own modules. The temptation is to include the thing you think they should know that they didn't ask about — sometimes valid, but the default should be to teach what they asked, because that's what they'll show up for and value.
Keep the audience's phrasing for titles. "Am I going to die?" beats "Safety." The emotional language they used is a gift; don't sand it into textbook register.
An extracted curriculum: its modules are the recurring questions from real responses, in the audience's language, ordered by what they actually care about. A stranger reading the module list would recognize their own questions.
An imposed curriculum: its modules reflect what the expert thinks matters, in expert language, possibly answering questions no one asked. It may be more "correct" and less wanted.
The test is whether the module titles sound like the audience or like the expert. If they sound like a textbook, you've imposed rather than extracted.
Self-reported with the potato gun example, no data.6 The method (build the curriculum from real audience questions) is sound and matches customer-development practice; the specific "8–10 from 100" figures are Brunson's.
Tension: extracting the curriculum from wants can systematically omit what the audience needs but didn't think to ask. The potato gun buyers asked about pipes and propellants; if there were a non-obvious safety factor no one knew to ask about, the extracted curriculum would miss it. Teaching only what's asked can leave dangerous gaps precisely where the audience doesn't know what it doesn't know.
Open question: how does an operator honor the extracted curriculum (what they asked) while still including the crucial thing they didn't know to ask — without sliding back into the expert-imposes-everything failure the Ask Campaign was meant to fix?
Convergence with the Ask Campaign is total — this is its direct downstream product. The 100 responses generate the curriculum here and the copy in the rewrite-questions-as-bullets page. Together they show the Ask Campaign feeding both the product structure and the marketing.
The tension with the vault's teaching corpus is the recurring want/need one. Good teaching sometimes means covering what students don't know to ask for; pure extraction covers only the asked. Brunson's method is strong on relevance and weak on the non-obvious essential, and the honest operator supplements the extracted curriculum with the few things expertise says must be included regardless.
To The Ask Campaign. The Ask Campaign gathers the 100 responses; this page turns them into curriculum. They're two halves of one motion — research and product design fused.
Held together: the Ask Campaign and the extracted curriculum collapse the usual gap between market research and product development — the research output is the product outline, with no translation step. Normally research informs a product the team then designs; here the recurring questions become the modules directly. Neither page states it: the method's real efficiency is eliminating the design phase, because the audience already designed the curriculum by asking.
To How False Beliefs Form. The questions people ask encode what they believe and fear — "Am I going to die?" reveals a belief-structure around danger. The extracted curriculum therefore also maps the audience's beliefs and anxieties.
The insight neither reaches alone: the recurring questions are a map of the audience's beliefs and fears, so the extracted curriculum is simultaneously a lesson plan and a psychological profile. The false-belief page explains how beliefs form; the extracted questions reveal which ones this specific audience holds. This means the same 100 responses that structure the course also feed the belief-breaking machinery downstream — the curriculum teaches the topics and the profile targets the objections.
Sharpest implication. The extracted curriculum can't be wrong about what the audience wants to know, because the audience wrote it by asking — and it arrives in their own language, with their own emotional weight intact ("Am I going to die?" becomes a module). It collapses the gap between research and product design into a single motion. But extraction has a blind spot exactly where the audience does: it covers only what people knew to ask, and can omit the non-obvious essential the audience didn't know it needed — which is where the expert's judgment has to supplement the method without sliding back into imposing everything.
Generative questions.
How does an operator include the crucial thing the audience didn't know to ask, while still honoring the extracted curriculum — where's the line between necessary supplement and expert overreach?
If the recurring questions are a psychological profile as much as a syllabus, does building the curriculum from them mean the course is optimized for the audience's anxieties as much as its learning — and is that good teaching or good marketing?