Guide

Test an idea with a landing page before building it

The “fake door” method: measure real demand with a page, not a product.

Updated July 22, 2026 · 6 min read

The best way to find out whether people want your product is to ask them to commit — before it even exists. That is the principle of the “fake door” (or smoke test): a simple landing page that presents the offer and invites people to sign up or pre-order. No code, no product. Just a full-scale demand test.

Why a landing page BEFORE the product

A survey tells you what people say they want. A landing page tells you what they do when asked for their email or their card. That is a huge difference: leaving your address is a micro-commitment, far more reliable than a polite “yes, I'd be interested”. You measure real intent for a few euros of ads — not weeks of development.

The 5 elements of a landing page that converts

A clear promise in the headline: the main benefit, in one sentence, no jargon.
A subheading that says who it is for and how it works.
3 concrete benefits (not features): what the person gains.
One reassurance element: launch price, “no commitment”, or some social proof, however modest.
A single call to action: email capture or pre-order. Not ten buttons — one.

Stay honest: state that the product is “in preparation”. A dishonest fake door burns your credibility. An honest fake door (“join the waiting list”) measures interest without deceiving anyone.

What conversion rate counts as a real signal?

It all depends on your traffic, but here are some useful benchmarks:

Email sign-up rate: below 5–10% of targeted visitors, the hook or the offer is not landing.
Pre-orders / reservations: even a handful of people willing to pay (or to reserve) is worth a thousand “likes”.
The traffic source matters: 50 sign-ups from a qualified audience beat 500 from an untargeted viral share.

Simple rule: if you cannot convince 100 targeted visitors to leave their email, you will not convince 1,000 to pay.

How to build a test landing page without coding

You do not need to develop anything. A no-code builder, a template, or directly the landing page generated by Sondari: describe your idea, get a ready-to-use waiting list / pre-order page, and send some targeted traffic to it (communities, ads, your network). You can even export the landing page as HTML to host it wherever you like.

The mistakes to avoid on a test landing page

Too much untargeted traffic. A thousand curious visitors give you a false negative. Aim for the right people.
A vague promise. If the visitor does not understand the offer in 5 seconds, they leave.
No measurement. Without tracking sign-ups/clicks, you know nothing. Put a minimum of tracking in place.
Giving up too fast. Test 2–3 hooks before concluding the idea is dead.

Building a test whose result is readable

Before sending traffic, write down your hypothesis: “out of 200 visitors matching profile X, at least 20 will leave their email to get Y”. Pick a single traffic source for the first test and add UTM parameters to every link. That way you will know whether a conversion comes from the promise or simply from friends already sold on your idea.

Qualified visitor: they belong to the defined segment and genuinely have the problem.
Primary conversion: a single action, decided before launch.
Threshold: a minimum number of visitors and a rate that triggers continue, change or stop.
Duration: long enough to cover the variation between weekdays and weekends.

Example of a small-budget landing page test

For an automation course aimed at estate agents, create two hooks: one promises to cut admin data entry, the other to speed up lead follow-up. Send each version to 100 visitors from the same channel and keep the same price, the same visual and the same form. If you change everything at once, you will not know what explains the gap.

After sign-up, ask one optional question: “which task would you want to automate first?”. The answers give you the vocabulary and the priorities of the future product. Then contact a few sign-ups to check their context, without selling them a solution that is already set in stone.

Respecting visitors and their data

A validation landing page must clearly state that the offer is in preparation, explain what the collected address will be used for, and allow people to unsubscribe. Do not fake testimonials, stock levels or countdowns. If you take a pre-order, state the estimated date, the refund terms and what happens if the project never ships. An honest test produces more reliable data and avoids destroying the trust of your first group of users.

Before you even build the landing page, make sure the demand exists: see our guide on how to validate a business idea.

Frequently asked questions

How do you test an idea with a landing page?

The principle is the “fake door”: publish a page describing the product as if it already existed, with a single call to action (leave an email, book a slot, pre-order), then send targeted traffic to it. What you measure is not the number of visitors but the conversion rate: the share of people who, having understood the offer, give something in return. It is the cheapest test that produces real commitment.

How many sign-ups mean the idea is validated?

The absolute number matters less than the rate and the threshold you set BEFOREHAND. Decide before the test what counts as a go (say 5% conversion on 200 qualified visitors), then hold yourself to it. A simple rule: a few hundred targeted visitors is enough to tell a strong signal from silence. And a sign-up is worth far less than a payment, however small.

Do you need to code to build a test landing page?

No. A landing page generator, a no-code builder or even a simple form is enough for a demand test. Coding it yourself at this stage is counterproductive: the whole point of the method is to spend as little time as possible before getting the answer. Sondari generates the landing page directly from the analysed idea and hosts it to collect the emails.

Is it dishonest to promote a product that does not exist yet?

It becomes dishonest if you take money for a product you do not intend to deliver. An honest pre-launch page states its status clearly — “in development”, “waiting list”, “coming soon” — and either takes no payment, or refunds in full if the project never ships. Framed that way, the fake door is a standard and transparent practice.

How long should you run the test?

Long enough to reach the visitor volume you decided on at the start, typically one to two weeks. Stopping as soon as the numbers look good is the most common bias: fix the duration and the threshold before launching, then read the result as it is, including when it disappoints.

From theory to a verdict

Sondari does all of this automatically: real demand, trend, competition and purchase intent cross-referenced in 30 seconds.

⚡ Crash-test my idea