build-before-they-ask — quality + safety report

---
name: build-before-they-ask
description: Force product-validation debates through Musk's "radically new products don't show up in surveys" filter — when the tech is genuinely new, users literally cannot tell you they want it, so customer-research signal will read negative and that negative reading is meaningless. Use this skill when the user is doing product validation, debating customer interviews, weighing a focus-group result, second-guessing a bold roadmap, hearing "no one is asking for this", or saying "should we do more customer research". Also fires when a team is about to kill a radical bet because the demand signal is quiet. The 1948 TV survey said 96% would never buy one. Tesla's early customers never asked for an electric car. Evidence accrues after you ship, not before. Trigger eagerly even when the user does not name Musk or the framework.
---

# Build Before They Ask

> "When you're building a radically new product, people don't know they want it yet, because it's just not in their scope."
> — Elon Musk, *The Book of Elon* (Chapter: Start Before the World Is Ready)

## What this skill captures

Surveys, interviews, and focus groups are useful instruments only when the respondent has the concept in their head. For radically new tech, the concept is not yet in scope — so the signal you get back is not "no demand," it is "no vocabulary." Musk's example is brutal: in 1946–1948, 96% of Americans said they would never buy a TV. At Tesla's start, Musk says he "heard zero times" that anyone wanted an electric car. The correct response is not more research — it is to ship, because *"as you make progress, evidence will build. More and more people will start to believe."*

The value to the user: stop letting null demand signal kill a bet that null signal can't actually evaluate. Decide on physics and conviction, not on a survey of people who can't yet picture the product.

## When to use this skill

- A founder is about to delay a launch because "users say they don't want it."
- A PM is running yet another customer-interview round to "validate" a category-defining feature.
- A board demands proof of demand for something no competitor has shipped.
- A team is sizing TAM by surveying buyers of a product that does not exist yet.
- An exec is using a focus-group result as a kill criterion for a radical bet.
- A founder is losing nerve because "no one believes in us yet."

## The how-to

1. **Classify the product first: incremental or radical?** If it is a faster horse, customer research is valid signal. If it is a car, the survey will lie to you.
   > "When you're building a radically new product, people don't know they want it yet, because it's just not in their scope."
   > — *The Book of Elon*
   Run this check explicitly before you touch any demand data.

2. **Discard null-demand signal on radical products.** Treat "no one is asking for this" as a tautology, not as evidence.
   > "In the beginning of Tesla, no one told us they wanted an electric car…I heard that zero times."
   > — *The Book of Elon*
   The absence of unsolicited demand for a category-defining product proves nothing about whether the product should exist.

3. **Decide on conviction and physics, not consensus.** Ask whether the thing should exist and whether the laws of physics allow it. If both, the survey is irrelevant.
   > "We say the things we believe, even when sometimes those things we believe are delusional."
   > — *The Book of Elon*
   Conviction in the face of negative early signal is the cost of admission for radical work.

4. **Stop seeking encouragement. Start shipping.** Encouragement is a lagging indicator of progress, not a prerequisite for it.
   > "If you need encouragement, don't start a company."
   > — *The Book of Elon*
   If the user is pattern-matching on "I need someone to believe in me first," name it and reject it.

5. **Generate evidence by shipping, not by asking.** The signal you actually need comes after the product exists in the world.
   > "Getting people to believe in you and what you're doing is important. In the beginning, very few people will believe in you or what you're doing. Over time, as you make progress, evidence will build. More and more people will start to believe."
   > — *The Book of Elon*
   Replace the next round of interviews with the next milestone that puts the product in front of real users.

6. **Run cheap, in-market evidence loops instead of hypothetical ones.** Prototype, preorder, alpha, demo day — anything where users react to the thing, not to a description of the thing.
   > "When you're building a radically new product, people don't know they want it yet, because it's just not in their scope."
   > — *The Book of Elon*
   People cannot describe demand for what they cannot picture; they can react to what they can touch.

## Common failure modes

- **Treating "no one asked for it" as a kill signal.** Musk heard it zero times for Tesla. This is the default state of radical products, not a verdict.
- **Running more interviews to break a tie.** If the first round returned null, the tenth round will too — the respondents still do not have the concept in scope.
- **Waiting for belief before building.** *"If you need encouragement, don't start a company."* Belief is downstream of shipping, not upstream.
- **Confusing a faster-horse product with a radical one.** Incremental products *do* show up in surveys. Misclassifying yours as radical lets you ignore real negative signal.
- **Confusing "delusional conviction" with "any conviction." ** Musk admits the beliefs may be delusional — the discipline is to act on them anyway *and* update fast once shipped evidence comes in.

## When NOT to use this skill

- The product is genuinely incremental — a better version of something users already buy. Their stated preferences are valid signal.
- The product is in a regulated, safety-critical domain (medical, aviation, finance) where shipping-then-learning carries catastrophic downside.
- You have already shipped and *real* in-market evidence is coming back negative — that is signal, not noise, and this skill does not license ignoring it.
- The user is using "radical product" as cover to skip basic feasibility, distribution, or unit-economics work.

## Source

The Book of Elon by Eric Jorgenson (2026, Scribe Media). Chapter: "Start Before the World Is Ready" (in "Living a Purposeful Life").