Validação
MVP
Produto
Startup
Lean
Pesquisa

Product Validation: How to Test Ideas Before Building

Most products fail because they solve problems that don't exist or that no one pays to solve. Validation is the process of testing hypotheses before investing in full development. This guide presents proven methods.

Why Validate

Reduce Risk

Discover problems before spending too much.

Save

Development is expensive. Validation is cheap.

Learn

Understand the market before committing.

Increase Chances

Validated products are more successful.

What to Validate

Problem

Does the problem exist? Is it important?

Solution

Does your solution solve the problem?

Market

Are there enough people with this problem?

Willingness to Pay

Would they pay for the solution?

Channel

Can you reach these customers?

Product Hypotheses

Problem Hypothesis

"[Persona] has [problem] when [context]."

Value Hypothesis

"[Solution] solves [problem] because [reason]."

Growth Hypothesis

"We were able to acquire [persona] through [channel]."

Document

Write hypotheses explicitly.

Validation Methods

Interviews

Deep conversations about problems.

Surveys

Quantitative data at scale.

Landing Page

Test interest with a simple page.

Concierge

Manually do what the product would do.

Fake Door

Measure interest before building.

Prototype

Usability testing without code.

MVP

Actual minimum version of the product.

Problem Interviews

Objective

Understand the problem, don't sell the solution.

Questions

"Tell me about the last time you..." "What do you do today to resolve it?" "What is most difficult about this process?"

Don'ts

Don't describe your solution. Don't ask hypothetical questions.

Quantity

5-10 interviews reveal patterns.

Landing Page Test

Elements

Value proposition, CTA for interest.

Metrics

Visitors, sign-ups, conversion rate.

Traffic

Ads to test quickly.

Validation

Do people sign up? Do they pay deposit?

Concierge MVP

Concept

Deliver manually what you would automate.

Benefits

Learn before building system.

Examples

Recommendations made by human, not algorithm.

Limitation

It doesn't scale, but it validates.

Fake Door Test

Concept

Button for feature that does not exist.

Measurement

How many click? They show interest.

Ethics

Explain that "coming soon" if they click.

Usage

Validate demand for a specific feature.

Testable Prototype

Fidelity

Enough to feel real.

Tests

Observe users using it.

###Feedback

Qualitative about experience.

Iteration

Improve based on learning.

MVP (Minimum Viable Product)

Definition

Minimal version that allows learning.

Objective

Test value hypothesis.

It's not

Bad or incomplete product.

Yeah

Functional product with minimal scope.

Validation Metrics

Qualitative

Feedback, insights, patterns.

Quantitative

Rates, conversions, numbers.

Signal

What indicates that it has been validated?

Threshold

Which number means success?

When to Pivot

Signs

Hypotheses repeatedly invalidated.

Pivot Types

Problem, solution, segment, channel.

Decision

Data + informed intuition.

Tools

Search

Typeform, Google Forms.

Landing Pages

Carrd, Webflow, Unbounce.

Prototypes

Figma, Marvel.

Analytics

Mixpanel, Google Analytics.

Common Errors

Confirmation Bias

Hear only what you want.

Leading Questions

Induce responses.

Building Too Much

MVP too big.

Ignoring Data

Continue even with negative signs.

Friends & Family

They are not real customers.

Conclusion

Validation is an investment that reduces risk. Test hypotheses early, listen with humility, and be willing to change. Validated products have much greater chances of success.

##FAQs

1) How long does it take to validate? Weeks, not months. Fast cycles.

2) Can I validate without code? Yes. Many methods do not require code.

3) How many interviews are enough? 5-10 for standards. More for statistics.

4) If people said they would buy it? Saying no is doing. Test real payment if possible.

5) When to stop validating and building? When you have enough confidence in the main hypotheses.

Also read