How to Validate a Digital Product Before You Build It

Three tests that cost almost nothing, what each one actually proves, and why asking people whether they would buy it tells you nothing.

27 September 20266 min read

Validation has a reputation for being heavy — surveys, landing pages, audience research, weeks of waiting. Most of that measures the wrong thing, slowly.

There is a much smaller version. It costs nothing, takes a few days, and answers the only question that matters: will a stranger give you money for this? Not a friend. Not a follower being nice. A stranger, with a card.

The short version

Asking "would you buy this?" produces a yes that means nothing. People are agreeable and imagining a purchase is free.
The only real signal is an attempt to pay, or a payment.
Test in order of cost: does a worse version already sell, will someone pre-buy it, can you deliver it by hand.
One stranger paying beats ten friends saying they would.
Validation tells you whether to build. It does not tell you whether it will be successful — those are different questions.

What validation is not

Three things get called validation and are not.

Asking people if they would buy it. Everyone says yes. Saying yes costs nothing, and people are socialised to be encouraging about a friend's project. A hypothetical purchase and a real one use different parts of the brain, and only one of them involves a card.

Asking your audience what they want. They will tell you what they want more of from you — the free stuff. That is not the same as what they will pay for, and the two are frequently opposites.

Building a landing page and counting email sign-ups. Closer, and still weak. An email address costs nothing to give. A meaningful sign-up rate tells you the message landed; it does not tell you the price works.

The failure mode this causes

The common story is: forty people said they would buy it, two did. That is not bad luck. It is what a yes-to-a-hypothetical is actually worth, and the forty conversations felt like evidence the whole time.

Test one: is someone already paying for something worse?

Free, instant, and it eliminates most ideas before you spend anything.

Find the closest existing product and read its one- and two-star reviews. Not to feel good about the competition being bad — to find out what the buyers wanted and did not get.

If you find a cluster of complaints about the same thing, you have a product brief written by the exact people who would buy yours. If the reviews are all positive, or there is nothing to review because nobody sells anything like it, stop. The second case is not an opportunity; it usually means the demand is not there.

What it proves: the category is real, and there is a specific gap in it.

What it does not prove: that you can reach those buyers, or that they will buy from you rather than the incumbent.

Test two: will a stranger pre-buy it?

This is the strongest cheap signal there is, and it is uncomfortable, which is why people skip it.

Put up a real page with a real price and a real way to pay, before the thing exists. Say plainly that it ships on a date. Then share it once, in one place where your actual buyers are.

You are looking for one number: did anybody try to pay? Not how many. Whether anyone did at all.

Describe the outcome, not the format

"This is a 40-page PDF" describes your work. "You will know what to charge for the next project without guessing" describes their outcome. Only one of those sells.

Put a real price on it

A free or "join the waitlist" version does not test the thing that matters. Price is the hardest part of the idea to get right and the easiest to test early.

Send it to one place, once

Not a launch. One post, to the audience most likely to have the problem. If the idea is real, one post is enough to produce a payment or two.

A pre-sale also does something a survey cannot: it tells you whether you can ask. Selling is a skill you have to be willing to practise, and finding out that you would rather not — before building the thing — saves months.

Test three: deliver it by hand

If pre-selling feels too exposed, this is the version with no audience required. Find three people with the problem. Sell them the outcome for a small amount. Then do the work manually — using your own spreadsheet, writing the document yourself, walking them through it on a call.

This is the slowest test and the most informative. You learn what people actually ask for, which parts they use, and which parts of your mental model were wrong. Two of the three will ask for something you had not thought of, and that something is usually the real product.

What it proves: people will pay, and you now know what for.

The risk: it does not scale, and it is easy to stay here. Manual delivery is a job. It is a way to learn the product, not a way to run one.

What "validated" means

A specific, unglamorous bar:

One person who is not your friend tried to give you money for it.

That is it. Not ten people. Not a hundred email addresses. One stranger, one payment, one instance of someone deciding your thing was worth more to them than the money.

Why one is enough

One payment proves the whole chain exists: there is a person, they have the problem, they can find you, they believe your promise, and the price is acceptable. Every additional payment is evidence about scale. The first one is evidence about whether the thing is real at all, and that is the question you are asking.

When to stop validating and build

Validation is easy to keep doing, because it feels like progress and carries no risk of the product being bad.

Stop when one of these is true:

Someone paid, and you know what to build first
Three people asked for the same thing and none of them paid — the demand is real, your offer is not
Nobody engaged at all, twice, from two different angles — the idea is not there, and that is a good outcome for the cost of a week

The third one is the one people avoid. Finding out an idea is dead in a week, for nothing, is the cheapest possible version of that news. The expensive version is finding out after four months of building — which is what happens when validation gets skipped because it feels like a distraction from the real work.

It is not a distraction from the real work. It is the only part of the process that tells you whether the real work is pointed at anything.

Where to look next

How to find a digital product idea — where the idea should come from in the first place.
How to price a digital product — the number you are testing, and how to choose it.
What online guides sell for — if the thing you are validating is a written guide, this is what the market pays.