How to use a landing page to validate a startup idea, and what to measure
One page, one promise, one action: the cheapest way to find out whether anyone wants what you plan to build.
What can a landing page tell you, and what can't it?
It can tell you whether your promise matters enough to strangers for them to act; it cannot tell you whether they will use the product every week. It tests demand, not habit. That is why it works best as a second step, after customer interviews and before any code; the wider method is in how to validate a startup idea.
- Smoke test
- An experiment that presents a product that does not exist yet in order to measure genuine interest: visitors see the offer and can act on it, while the product comes later. Done honestly, the page says plainly that the product is still being built.
An email address is a weak signal, a booked call is stronger, and picking a paid plan or placing a pre-order is the strongest signal a page can give you. Design the page around the signal you need, not the one that is easiest to collect.
How do you build a landing page to validate an idea?
From hypothesis to live page
Write the hypothesis before the page
Who has the problem, what you promise, which action you expect and what result would make you carry on. Set the number after seeing the data and you will justify any outcome.
Lead with the customer's outcome
Not the product name: the change it brings. One line underneath on how.
Three sections, not ten
The problem in the customer's words, how it works in three steps, what happens after sign-up. No feature lists to fill space.
One action only
A single form or button, repeated at the top and the bottom. Every extra link is another way to leave without answering your question.
Add a step to test price
To learn about willingness to pay, put a choice of priced plans between the pitch and the form. A click on a paid plan tells you something an email address never will.
Tracking in place before traffic
Visits, completed actions and the source of every visit. Check your local cookie rules before you switch any tracking on.
Is there a real example of validating with a landing page?
Yes, and it is still one of the clearest. Buffer's founder Joel Gascoigne described it in a 2011 post: the first version was a page explaining the idea and collecting emails, to find out whether people would even consider using the product. Once interest showed up, he added a pricing page between the pitch and the sign-up form, so each visitor made one extra click and some chose a paid plan.
Only then did he build the real, rough-edged product, and by his account the first paying customer arrived within four days of launch. The lesson is not speed but sequence: evidence of demand first, then of price, then code.
Where should the traffic come from?
From people who have the problem, not from anyone at all. The best sources are the places your audience already gathers: industry communities, newsletters, professional groups, the people you interviewed. Paid ads help when you want a larger and less friendly sample; how to run a small budget is covered in Google Ads for startups.
Which numbers should you look at?
Look at the share of visitors who take the action, split by traffic source, and compare it with the target you wrote down beforehand. A blended figure mixes enthusiastic friends with indifferent strangers and tells you nothing about either.
| Signal | What it measures | How to read it |
|---|---|---|
| Time on page, scroll depth | Attention | Useful for fixing copy, not for deciding |
| Waitlist sign-up | Low-cost interest | Counts only when it comes from strangers in your target |
| Answer to a follow-up question | Interest with a little effort | The answers matter as much as the count: read them all |
| Booked call | Interest backed by time | A strong signal for B2B products |
| Click on a paid plan | Stated willingness to pay | Compare plans: it also tells you where to price |
| Pre-order or deposit | Real willingness to pay | The strongest signal; requires clear terms on timing and refunds |
6.6%
Treat it as a reference point, not a target. The figure covers pages from companies already in market, with every kind of traffic and offer; a validation page fed with cold traffic can sit below it and still give you a clear answer, provided the right people take the right action.
What can you do with the emails you collect?
Use them for the purpose stated on the form, such as announcing the launch, and nothing else without specific consent. In the UK the ICO's guidance on electronic mail marketing is blunt: you must not send marketing emails to individuals unless they specifically consented or are existing customers of a similar product. EU countries have their own versions: in Italy, for example, Article 130 of the Privacy Code requires consent for marketing emails. Say on the form what people will receive, link your privacy notice and ask separately for newsletter consent; our GDPR guide for startups covers the rest.
How do we prepare a validation landing page?
Once the landing page has done its job, the next step is not a full website but a minimum product, as explained in what is an MVP. The page stays useful as the seed of your future site, following the plan in startup website structure.
Frequently asked questions
How many visitors do I need to validate an idea?
There is no universal number: it depends on the size of the difference you are looking for and how targeted the traffic is. A few hundred visits from the right audience say more than thousands of random ones. Set your threshold first and stop when results per source stop moving.
Is a waitlist enough proof?
It is a first signal, not proof. Sign-ups cost people very little, so strengthen them with a follow-up question, a call or a pricing choice. The more an action costs the visitor, the more you can trust it.
Is it ethical to present a product that doesn't exist yet?
Yes, if you say so clearly: “we're building”, “coming soon”, “be among the first”. Taking money for something you will not deliver, or advertising offers that are unavailable, is not. Transparency also protects the value of the test.
Should I use a page builder or a custom page?
For a first test a builder is fine, as long as the page is fast and tracking is right. A custom page pays off when the landing page will become the base of your site or when performance matters for paid traffic. What counts is being able to change headline and offer quickly.
What if the landing page doesn't convert?
Check the traffic first: if visits do not come from your target audience, the test says nothing. Then change one thing at a time, starting with the headline promise. If the right audience ignores several different promises, that is valuable too: it just saved you a build.
Sources
- Joel Gascoigne (Buffer), Idea to Paying Customers in 7 Weeks: How We Did It (16 febbraio 2011) (19 October 2026)
- Unbounce, Conversion Benchmark Report 2024 (19 October 2026)
- Google Ads Help, Misrepresentation policy (Unavailable offers) (19 October 2026)
- Normattiva, D.Lgs. 30 giugno 2003 n. 196 (Codice privacy), art. 130 (19 October 2026)
- ICO, Guide to PECR: electronic mail marketing (19 October 2026)
- Garante per la protezione dei dati personali, Linee guida cookie e altri strumenti di tracciamento (10 giugno 2021) (19 October 2026)
Figures and definitions checked on the dates shown.
