By Toon Timbermont in Product — 03 Jan 2026

5 Ways to Validate Demand Before You Build

Move from demand thinking to demand testing with these proven validation techniques, from surveys to behavioural experiments.

Photo by Franck / Unsplash

In my previous learning loop, I made the case for demand thinking over solution thinking. For asking "will people pay?" before "can we build it?" For channelling your team's (builder) energy toward validation first.

But I left you with a gap.

Knowing you should validate demand is one thing. Knowing how to do it—concretely, practically—is another. So, in this learning loop, I want to make things tangible. With five techniques I find particularly useful for validating demand—not a definitive list, but a practical starting point following a natural research progression: from qualitative discovery, through quantitative validation, to behavioural and solution testing.

Think of it as a validation journey. Each technique builds on the previous one, answering different questions with increasing confidence.

1. The 4 Forces Framework: Why Would They Switch?

A technique for Customer Interviews

Start with conversations. Before you can validate anything at scale, you need to understand why people would change their current behaviour. The 4 Forces framework from Jobs-to-be-Done theory maps the dynamics of every switching decision:

Forces pushing toward change:

Push of the current situation — What frustration creates the "struggling moment"?
Pull of the new solution — What's the vision of a better life?

Forces resisting change:

Anxiety of the new solution — What fears hold them back?
Allegiance to the present — What habits keep them with the status quo?

Conduct 10-12 interviews with people who recently switched. Map their timeline from first thought to final decision. Look for patterns across the forces.

The hidden gem: Often the biggest barrier isn't your product—it's the anxiety about change. Find this bottleneck, and your product will benefit a lot

You can find more info and a handy prompt in the post From Assumptions to Insights: Customer Discovery That Works.

2. MaxDiff Analysis: What Matters Most?

A technique for Validation Surveys

Once your interviews have surfaced patterns and you have some early ideas about solutions, you need to validate them at scale. TIP: Don't fall into the confirmation bias, asking people if something is important... people rate everything as "important" to be polite.

MaxDiff (Maximum Difference Scaling) addresses this and forces trade-offs. Respondents see sets of 4-5 options and must choose the most and least important. After cycling through multiple sets, you get a clear ranking of what actually matters.

When to use it: After qualitative interviews have surfaced potential directions. When you want to understand which features are really influencing your target user’s choice.

Limitation: MaxDiff measures stated preferences, not behaviour. Always pair with behavioural validation.

3. Fake Door Tests: Would They Click?

Also called Feature Stub, Smoke Test, or Painted Door

Interviews and surveys measure what people say. Fake door tests measure what people actually do.

A fake door presents users with what appears to be a real feature—a button, a link, a landing page. When they click, they see a message explaining that the feature is coming soon, often with an option to sign up for updates.

Famous examples:

Buffer launched with just a landing page and a "Plans & Pricing" button. Clicks validated demand before any code was written.
Dropbox posted an explainer video on a placeholder page. 70,000+ waitlist signups overnight.

When to use it: When you want quick, cheap validation. When you can access real traffic. When you need behavioural data, not opinions.

Important caveat: Be transparent. "Coming soon" messaging prevents users from feeling tricked.

4. Wizard of Oz: Would They Use It?

Sometimes you need to test whether your solution actually works—not just whether people are interested. A Wizard of Oz presents a fully functioning product experience, but behind the scenes, humans do the work manually instead of automated systems.

The user doesn't know there's a "wizard" behind the curtain. They experience what feels like a real product.

Famous example: Zappos started by photographing shoes at local stores and only buying them when customers ordered. The website looked like real e-commerce, but inventory was manually sourced per order.

When to use it: When automation would be expensive to build. When you need to validate the experience, not just interest.

Key insight: The goal isn't to run a Wizard of Oz forever—it's to learn whether automation is worth the investment.

5. Concierge: Would They Pay?

The Concierge is similar to Wizard of Oz, with one crucial difference: the customer knows they're getting a manual, high-touch service. And critically—they pay for it.

This is the highest-confidence validation because it tests willingness to pay with actual transactions. If people open their wallets, you have demand. If they don't, no amount of "interest" surveys matter.

Famous example: Food on the Table started without any software. The founder personally created customised meal plans and shopping lists for individual families—for a fee.

When to use it: When you need to validate unit economics. When you want the highest confidence before major investment.

The power of manual: Yes, it doesn't scale. That's the point. You're not trying to scale yet—you're trying to learn whether scaling is worth it.

The Validation Spectrum

Each technique serves a different purpose and provides different confidence levels:

Technique	What it Tests	Confidence
4 Forces	Motivations	Medium
MaxDiff	Preferences	Low-Medium
Fake Door	Interest (actions)	Medium
Wizard of Oz	Solution fit	Medium-High
Concierge	Willingness to pay	High

The art isn't picking one technique—it's knowing where you are in your validation journey and which questions you need answered.

From Thinking to Testing

Last learning loop was about mindset—shifting from solution thinking to demand thinking. This loop is about method—the concrete techniques that turn demand questions into validated answers.

Which (other) technique would give you the confidence you need for your riskiest assumption?