Choosing Opportunities and Validating Demand

Learning goal: Size opportunity candidates bottom-up, confirm demand before building with Mom Test interviews, fake doors, pre-sales and waitlists, and decide against pass criteria set in advance.

As of: 2026-09-29. Market sizes and threshold numbers are all assumptions; fill in real values with your own research.

Key Concepts

Principles

Where Opportunities Come From

Market Size for Small Products: Bottom-up

For a small product, a top-down estimate such as "1% of the global design software market" is useless. Build up from below: who, how many, and how much they pay.

TAM = potential customers with the problem × annual willingness to pay SAM = the part of TAM you can reach with your language, platform and payment methods SOM = the share of SAM you can realistically win in the first one to two years

Example: a hypothetical creative tool (all numbers are assumptions)

• Bad example: "TAM is KRW 1 billion, so taking just 1% gives KRW 10 million a year."
• Good example: "SOM is about KRW 330,000 a month, roughly 10% of break-even (KRW 3,500,000 a month) from this one product. Treat it as a side product that covers the KRW 500,000 fixed cost, or proceed only if there is a way to widen SAM."

Validating Demand Before Building: The Evidence Ladder

Actions are stronger evidence than words, and money is stronger than actions.

flowchart TD
    S1[Compliments and opinions] --> S2[Email sign-up]
    S2 --> S3[Waitlist reply or return visit]
    S3 --> S4[Click on the buy button of a price page]
    S4 --> S5[Pre-payment or deposit]
    S5 --> T{Is the highest signal reached at or above the preset threshold}
    T -->|yes| GO[Move to a build slot]
    T -->|no| RE[Revise the message once and measure again]
    RE -->|short again| AR[Archive]

• Fake door: Tell anyone who presses the button right away that "this is not ready yet and we will notify you at launch." Do not take money or pretend it exists.
• Pre-sale: Real payment makes it the strongest signal. It also creates e-commerce obligations such as refund terms, a delivery date and mail-order sales registration. Read the Business section's "Registration · Mail-Order Sales" and "Terms · Privacy · E-Commerce" first.
• Waitlist: Collecting emails requires consent to personal data collection and a privacy policy. Filter signal quality once more with reply and return-visit rates after sign-up.
• Steam wishlist: A strong advance signal for games. How to read it is covered in 07.

Mom Test Interviews

In The Mom Test (2013), Rob Fitzpatrick set out three rules for asking questions even your mom cannot lie to you about: talk about their life instead of your idea, ask about specifics in the past instead of opinions about the future, and talk less and listen more.

Set Signal Thresholds First

If you set the criteria after the experiment, any result reads as "promising". Write one line before the experiment.

Hypothesis / metric / period / pass threshold / action if passed / action if missed

Example (assumption): "Productivity app P solves freelancers' schedule-sorting problem / waitlist sign-ups per landing-page visit / 2 weeks / at least 10% of 300 visitors / move to a build slot / revise the message once and measure again; if short again, archive."

Applied: The Example Studio

This lays out the products placed in validation slots in 02 and the first validation method for each other type. All thresholds are assumptions, run within the monthly validation budget of 24 hours.

The technical documentation site (C2) already has articles and is past this validation step. So the traffic math in 05 comes before "demand validation", and its gates use the same unit (monthly pageviews) in 09 and 10. If visitor numbers fall far short of what the target RPM math requires, consider other models (04) alongside ads.

Going Deeper

Conversion rates wobble in small samples. The standard error of a conversion rate p is √(p × (1 − p) ÷ n) (a normal approximation, and a rough one for small samples).

A 10% seen from 50 visitors might really be 2% or 18%. So set a minimum sample size in advance along with the threshold. Not having enough traffic to fill the sample is itself an important signal: "there is no acquisition channel".

Common Misconceptions

• "We got lots of compliments, so there is demand" — Compliments sit at the bottom of the evidence ladder.
• "A big TAM means a good opportunity" — A solo studio's problem is SOM and the path to reach it.
• "No competitors means opportunity" — There may be no competitors because there is no demand. First check whether anyone already spends money.
• "A fake door deceives users" — Telling the truth immediately and taking no money makes it an honest experiment. The moment you take payment, it is a pre-sale with legal obligations.
• "If validation is ambiguous, build a bit more" — Ambiguous means missed. Revise the message once, measure again, and stop if it misses again.

Self-Check Questions

1. Write a bottom-up TAM/SAM/SOM for one current candidate. What percentage of break-even is the SOM?
2. Find one question from a recent customer conversation that broke the Mom Test and rewrite it as a good question.
3. What is the highest rung of the evidence ladder your product has reached so far?
4. With 100 visitors and a 10% sign-up rate, compute the rough 95% range.
5. What are two legal obligations to check before starting a pre-sale?

References

• The Mom Test — Rob Fitzpatrick (Goodreads) (2013, accessed 2026-09-28)
• The Mom Test by Rob Fitzpatrick — mtlynch.io book report (accessed 2026-09-28)
• Fake Door Testing: What It Is and How to Run One — Learning Loop (accessed 2026-09-28)