How to validate a startup idea without writing a single line of code

Building is the most expensive way to test an idea. Here are five patterns that buy real information cheaper, faster, and often more honestly. Pick the one that matches your riskiest assumption.

1. The landing-page test (validates: do people care?)

Build a one-page site that describes the product as if it exists. Add a single CTA: "Get early access" → email capture.

Buy $50-200 of targeted ads. Measure: email signup rate. If less than 3% of targeted visitors give you an email, the message or the audience is wrong — fix that before building.

2. The fake-door test (validates: which feature do people actually want?)

You have three feature ideas. Build a landing page for each, run identical ads to each. The one with the highest click-through rate is the feature people want — start with that one.

Cost: 2 hours of work, $30 in ads. Outcome: you stop building the feature you assumed was best.

3. The concierge MVP (validates: can you actually deliver value?)

Promise the outcome. Deliver it manually. The "AI" is you with a spreadsheet and an inbox.

A friend of mine "launched" an AI-powered hiring tool and spent the first month manually reading every CV and writing the summary by hand. By the time it was worth automating, he knew exactly what to automate because he'd done the work 200 times.

4. The Notion + Stripe stack (validates: will people pay?)

A Notion page with a Stripe payment link. Customer pays, you email them the deliverable.

This works astonishingly well for digital products (templates, guides, audits, consultations). I have seen multiple $30k MNR products that ran on this stack for a year before anyone wrote real code.

5. The Telegram bot pilot (validates: will people use it daily?)

Build a 2-day Telegram bot that asks for one input and returns one output, hardcoded. Share with 20 target users.

Daily active usage tells you if the workflow is worth the build. Zero engagement tells you to pivot before spending three weeks on a polished version.

What to measure (and what to ignore)

Track:

• Conversion to action (email, click, payment) — the only number that proves intent.
• Repeat behavior in week 2 — proves the value, not the curiosity.
• Unprompted feedback — what users say *before* you ask is more honest than survey answers.

Ignore:

• Total visits — vanity if nobody acts.
• Time on page — could mean engaged or could mean confused.
• Survey responses about willingness to pay — people lie. A $10 payment proves more than 100 survey "yes" answers.

The goal of validation is not to prove your idea works. It is to find out how to change it so it can.

Spend a week on one of these patterns before writing a line of code. The information you buy will reshape what you build — and that reshape is exactly the value of validation.