Validating Demand Before You Build

You cannot prove your product will make money before you build it. You can prove that money already flows to products shaped like yours, for your customer, and you can make strangers pay you before a line of code exists. Everything else is theatre.

The concept

There's a comforting fiction that you can "validate" an idea cheaply: put up a landing page, run $100 of ads, count the emails, and now you know. You don't. A landing-page signup validates that your headline is good. It says nothing about whether anyone will pay $50/year for a product that doesn't exist.

Validation is not binary and it is not certainty. It's a ladder of evidence, and your job is to climb as high as you can before committing months.

The evidence ladder

   STRENGTH
   ▲
   │  ██████████  Your own recurring revenue        ← only exists post-launch
   │  █████████   Paying pre-orders / deposits      ← ⭐ strongest pre-build
   │  ████████    LOIs / signed pilots (B2B)        ← ⭐ strongest pre-build, B2B
   │  ███████     Competitor revenue, YOUR SHAPE    ← ⭐ free, one afternoon
   │  ██████      Competitor revenue, your category
   │  █████       Demographic data on your customer
   │  ████        Workarounds people already built  ← from interviews (Ch 04)
   │  ███         Category growth rate
   │  ██          Free waitlist signups             ← validates the headline only
   │  █           Social engagement / "great idea!" ← validates nothing
   ▼

Everything from "competitor revenue, your shape" upward is obtainable in one afternoon, for free, before you write any code. Most founders skip those rungs entirely and go straight to a landing page, which sits three rungs lower.

The structural comp: the highest-value free evidence

This is the single most useful concept in the chapter.

You are looking for a company that is your shape with a different topic, serving your customer, at your price point. Not a competitor, a structural analogue.

• Building a 90-day habit course for young men? Find any 90-day paid program for young men that works. Topic irrelevant.
• Building a B2B tool for dental practices? Find a comparable tool for veterinary practices.
• Building an AI writing tool for lawyers? Find an AI tool for accountants.

If the shape earns money for that customer at that price, the shape is proven. Your remaining risk is execution and distribution, which is a very different, much smaller risk than "does anyone want this at all."

Why this beats topical comps: topic tells you about interest. Shape tells you about economics, willingness to pay, retention profile, price tolerance, sales motion.

Demographic reversals

When you find a category that inverts its usual demographic skew, you've found a real, structurally-caused sub-population rather than a slice of a spreadsheet. Personal development normally skews female; if a sub-category is 60% male, something real is causing that, and real causes make real markets.

Look for reversals. They're strong signals and they're free to find.

What "validated" honestly means

You are never validated. You are less wrong. The purpose is to convert "I think people want this" into "here is why I believe people want this, and here is the specific thing that would prove me wrong."

📐 Best practice

Write a GO/NO-GO memo with an actual question at the top. Not a summary, a question the document is capable of answering "no" to. If every section conveniently agrees with you, you wrote a justification, not an analysis.

Find three revenue comps, at least one structural. Table: product, positioning, downloads/customers, revenue estimate, pricing.

Take money before you build. The strongest pre-build evidence that exists:

• Consumer: a founding-member lifetime deal or a discounted pre-order at 40-60% off, capped and honest ("first 500")
• B2B: a signed LOI, a paid pilot, or a deposit against a delivery date

Ten people paying $20 tells you more than a thousand free emails. It validates price, not just interest, and it funds your first costs.

Look past every statistic to the behaviour it implies. "$5.8B market growing 16%" changes nothing. "Men use this instead of therapy because it doesn't feel like weakness" changes your copy, your visual identity, your onboarding, and your store listing. Hunt for the second kind.

Write the risks section honestly, including where your own evidence is weak. A memo with no risks section is marketing.

Name how each comp got its users. If a comparable company grew on $500k of TikTok ads and your plan is "I'll post sometimes," you are not comparable no matter how similar the product is.

Re-read the memo at every milestone. Ten minutes. Products drift from their positioning silently and nothing else catches it.

💀 Common mistakes

Validating the topic instead of the shape. "Stoicism is popular!" is not evidence that a Stoicism course sells.

Cherry-picking winners. Every comp is a success story because you searched for successes. Deliberately search "[category] shut down" and "why I killed my app." Balance the table.

Treating free signups as demand. A 5% waitlist-to-paid conversion is common and often optimistic. 500 free emails is not 500 customers.

Ignoring survivorship bias. For every Quittr or Superhuman you've heard of, dozens of identical products made nothing. Comps prove the ceiling exists, never that you'll reach it. The correct reading is always "this shape can work", never "this will work for me."

Writing the memo to justify a decision already made. Test: have someone else, or a fresh AI session with no context, write the NO case. If it's weak, you're fine. If it's strong and you're irritated, you've learned something.

Confusing "no competitors" with opportunity. Usually means no market.

Building the "smoke test" landing page and calling it done. A fake-door test tells you about your headline and your traffic source. It's rung two.

Letting the product drift from the validated positioning. You validated one thing and shipped another. Nobody notices for months.

The professional workflow

 1. WRITE THE QUESTION the memo must answer.
    It must be answerable "no".

 2. IS THE MARKET REAL AND GROWING?
    Cite sources. For every stat, write the BEHAVIOUR it implies.

 3. IS THE AUDIENCE WHO I THINK?
    Demographic data. Look specifically for REVERSALS of the
    category norm — they mark real sub-populations.

 4. DO PRODUCTS OF MY SHAPE MAKE MONEY WITH THIS CUSTOMER?
    ⭐ Table of 3+ comps, at least one STRUCTURAL.
    For each: how did they get their users? Can I replicate it?

 5. THE UNCOMFORTABLE STRATEGIC QUESTION
    Every idea has one (who to exclude, what to charge, what not
    to build). Argue both sides in writing. Resolve it.

 6. RISKS — including where my own evidence is weak

 7. ⭐ TAKE MONEY
    Consumer: founding lifetime / pre-order, capped, honest
    B2B: LOI, paid pilot, deposit
    Target: 10 payments, or a clear "no" in two weeks

 8. ONE-LINE VERDICT + DATE. Diarise a re-read.

Fake-door and smoke tests: done properly

If you can't take money yet, the next best thing is a fake door: a landing page with a real CTA that leads to "not ready yet, join the list." Rules that make it meaningful:

• Send real, targeted traffic (a relevant community, a small ad set), not friends
• Measure click-to-CTA rate, not visits
• Put the price on the page. A signup without a price validates nothing about willingness to pay
• Add a second step: "reserve your spot for $5, fully refundable", this is the single highest-signal thing you can do

Ethics: never take money for something you don't intend to build, and always disclose that it's pre-launch. A refundable deposit with a clear delivery window is honest. A fake checkout that charges people is fraud.

Tools, websites & costs

Realistic total: $0-250. Every source in a good comps table is free and public.

The validation-memo prompt

Write a GO/NO-GO validation memo for [IDEA] targeting [SPECIFIC CUSTOMER].
Open with the actual question. The memo must be able to conclude NO.

1. Is the market real and growing? Cite sources. For every statistic, state
   the BEHAVIOUR it implies — not just the number.
2. Is the audience really who I think? Find DEMOGRAPHIC data. Look
   specifically for REVERSALS of the category norm.
3. Do products with MY SHAPE make money with THIS customer? Table: product,
   positioning, downloads, revenue, pricing. Prioritise STRUCTURAL comps
   (same shape, different topic) over topical ones.
   For each comp, state HOW they acquired users and whether that is
   replicable by someone with no audience and no budget.
4. Deliberately find 2 FAILURES in this space and why they failed.
5. The uncomfortable strategic question: [X]. Argue both sides. Resolve it.
6. Risks — including where this memo's own evidence is weakest.

Tag every figure [VERIFIED — source] or [ESTIMATE]. End with a one-line verdict.

Alternatives & trade-offs

Concierge / Wizard-of-Oz, deliver the outcome manually before automating it. Extremely strong validation (people pay for the outcome), and it teaches you the workflow you're about to encode. Slow, unscalable, and the right first move for most B2B and service-shaped products.

Pre-selling a cohort, sell a first cohort at a discount with a start date. Common for courses and community products; forces a deadline and produces real customers.

Building in public, ship progress on social while building. Validates interest and builds an audience simultaneously. Works if you can sustain the posting cadence; otherwise it's a distraction with a deadline.

Just build it, legitimate when the build is genuinely a weekend, when you are unambiguously the customer, or when the market is too novel to research. Indefensible for a three-month build.

Trade-off: every rung up the ladder costs more time and gives more certainty. Match the rigour to the size of the bet. A two-week build needs rung 4. A six-month build needs rung 8 or 9.

Checklist

• I have a written GO/NO-GO memo with a question it could answer "no" to
• I have 3+ revenue comps, at least one structural (my shape, different topic)
• I know how each comp acquired users and whether I can replicate it
• I deliberately looked for failures in this space
• I have demographic evidence the audience exists at the scale I assume
• For every statistic I wrote down the behaviour it implies
• I identified and resolved the uncomfortable strategic question
• I have a risks section naming where my own evidence is weak
• Every figure is tagged [VERIFIED] or [ESTIMATE]
• ⭐ I have taken money from strangers, or know exactly why I can't yet
• The price was visible on anything I used to measure interest
• I have diarised a re-read of this memo at my next milestone

📓 Case Study: a memo that could have said no

Project: SOLIS. The validation work was written up as a dated memo opening with a question rather than a conclusion:

Question: Is there a real, paying market for a classical/stoicism-flavored discipline app, and is positioning toward men a strength or a limit?

Verdict: GO, the niche is real, growing, and male positioning is a strength, provided we lead male rather than exclude women.

The structural comp, this is the rung that mattered.

Quittr is the important row. It is SOLIS's shape with a different topic: a 90-day structured program, young male audience, progress tracking, subscription. Different subject entirely, and it cleared roughly $3M in year one.

That doesn't prove SOLIS will earn anything. It proves the shape earns money from this customer at this price, which reduces the open risk from "does this category exist" to "can I execute and distribute." Free, one afternoon.

The demographic reversal. Stoic Week surveys: 60% male (2019), 57% (2020), and the organisers explicitly noted this reverses the usual personal-development pattern. The inverse confirmed it: meditation app users skew ~65% women; Calm subscribers ~80% women. Two independent measurements pointing opposite directions and agreeing on the same conclusion. That's what good evidence looks like, not one big number, but triangulation.

Behaviour behind the statistic. The memo didn't stop at "Stoicism content grew ~400%." It recorded why: the growth was driven largely by men using it as a DIY mental-health tool without the perceived vulnerability of therapy. That single clause shaped the product's tone, its visual identity, and its store listing far more than any market-size figure could.

The uncomfortable question, answered in writing. Is targeting men a strength or a limit? Both sides were argued, differentiation and channel fit versus halving the addressable market and inviting review friction, and resolved as "male-coded, not male-gated." The aesthetic (emperors, legions, oil paintings) does the targeting; the copy stays universal. Aesthetics select an audience without a single exclusionary word.

⚠️ Deviation 1: no money was ever taken.

Best practice is to climb to the top of the ladder, take pre-orders. The product plan even specified the instrument: "Founding lifetime (launch only): $79.99, capped (e.g. first 500). Generates cash before the AI backend exists and creates urgency."

It was never built. The waitlist collected free emails only. So the project reached rung 7 (structural comps) and rung 2 (free signups), and skipped rungs 8-9 entirely.

The consequence is precise: after 24 days of building, the price was never tested. Nobody was ever asked to pay $49.99, or $79.99, or anything. The single most answerable question, will this customer pay this price?, remained open at the end of the build, and a half-day of work on day one would have started answering it while also funding the $99 developer account.

⚠️ Deviation 2: the memo was never re-read, and the product drifted.

The memo said male-coded, not male-gated. Seventeen days later the product had drifted: a phase named "The Man," a pitch line of "Build the mind of an emperor," and, briefly, a gender question in onboarding. The contradiction was noticed by accident on day 22 while building something else, and the gender question was cut.

Nothing in the process caught a 17-day drift from a validated position. A ten-minute re-read at each milestone would have.

🚩 Still unvalidated. The memo's own risks section contains the line that aged into the truth of the whole project: "the funnel converts traffic, but traffic remains the real fight." That warning was written on day 5. Twenty-four days later there was a finished app, a live site, a working backend, a payment system, and zero users. Writing the risk down is not the same as weighting it.

Lessons

1. You can't validate your product pre-build. You can validate your shape. Find your shape with a different topic, it's the strongest free evidence available.
2. Climb the ladder as high as your build size demands. A three-month build needs more than a landing page.
3. Take money. Ten $20 payments beat a thousand free emails, validate price, and fund your first costs.
4. Look past every statistic to the behaviour it implies.
5. Demographic reversals mark real audiences.
6. Write the memo so it can say no, with a risks section that names your weakest evidence.
7. Note how every comp got its users. If it was $500k of ads, you are not comparable.
8. Re-read your memo at milestones. Positioning drifts silently; nothing else catches it.
9. Writing a risk down doesn't mean you've weighted it. "Traffic is the real fight" was on paper from day 5 and under-weighted for the next 19.

Next: Chapter 06: Positioning and Product Strategy →