What is a smoke test?

Why is it called a smoke test?

The name comes from hardware: switch the device on and see whether it smokes. In product work, what you test is the product's shadow – its description, price and promise – before the product itself is built. In Swedish we call it a skuggtest, a shadow test.

Related names are fake door test, demand test, landing page test and pre-launch validation. The principle is the same.

How does a smoke test work?

1. Decide what you want to know, and which result would convince you. For example: at least 5% of visitors sign up.
2. Make a page that describes the product: what it does, for whom, and what it costs.
3. Send people who resemble your intended customers to it – through ads, a newsletter, a community or your own network.
4. Measure what they do: how many arrive, how many click, and how many leave their email address.
5. Compare with the line you drew beforehand. Above it: keep building. Below it: change the idea or let it go.

A typical smoke test runs for two to four weeks and needs a few hundred visitors to give a signal you can trust.

When is a smoke test the right tool?

When the question is whether anyone wants it, not whether it can be built. It suits new products and services, new audiences and markets, prices and packages, and ideas you are thinking of spending months on.

It suits less when the product can only be judged by using it, or when you cannot reach your intended customers at all. Interviews or a simple prototype are often better then.

Is it honest?

Yes, if the visitor is told the truth. Anyone who clicks should learn straight away that the product does not exist yet, and ideally be able to sign up to hear more. Never take payment for something that does not exist.

Most people like being in early. Someone who feels tricked does not – and gives you worse data too.

What counts as a good result?

It depends on how much you ask for. A click on a button is easy to give; an email address costs more, and a click on a price says the most.

As a rule of thumb: a few percent of visitors leaving their address is a clear signal for a new product. Most important is to set the line before you see the numbers, so you do not explain away a weak result afterwards.

Common questions

What is the difference between a smoke test and an MVP?

An MVP is a first version of the product that can be used. A smoke test comes before it: it shows whether an MVP is worth building at all.

How many visitors do I need?

A few hundred from the right audience is usually enough for a first signal. Fewer than a hundred rarely gives an answer you can trust.

Do I need to code?

No. With Skuggtest you write the text and choose a style; the page, the measuring and the collecting of email addresses are included.

May I collect email addresses?

Yes, if the visitor is told what the address is for and chooses to leave it. Skuggtest shows such a text by the form and keeps the addresses for you.