How to validate a SaaS idea without writing code
In short
Signups measure how well your page asks, not whether anyone wants the product. Ask what people would pay instead, give them a real option to refuse, and count how many name a price. Two founders with 7,800 and 6,264 registered users converted 0.23% and 1.1% of them.
Why doesn’t a landing page tell you if a SaaS idea works?
The standard advice is to put up a landing page, collect emails, and treat the signups as proof. It is cheap and it feels like evidence. The problem is what a signup costs the person giving it: an email address, and about four seconds. Nothing they would miss.
So the number you collect measures how good your landing page is at asking, not how much anyone wants the product. Those are different questions, and only one of them decides whether you have a business.
What do signup numbers actually convert at?
Two founders published their own figures, counting registered users against people who eventually paid:
| Registered users | Paying | Conversion |
|---|---|---|
| 7,800 | 18 | 0.23% |
| 6,264 | 71 | 1.1% |
The published benchmarks for waitlist-to-paid conversion generally quote 5% to 25%. These are between five and a hundred times worse. Both founders had what looked like traction — thousands of people — and neither had a business.
The gap is not because these two were unusually bad. It is because the benchmarks are written by people with something that worked, and the ones that did not work do not publish. If you are using 20% to plan with, you are planning against a number that survived selection.
How do you validate a SaaS idea without writing code?
Ask for the one thing that costs something to say. Not “would you use this” — everyone says yes to that, because saying yes is free and being encouraging is polite. Ask what they would pay, show them a real number, and let them decline.
The mechanics, in order:
- Describe the product as if it exists.One page, plain language, the specific problem it solves. No roadmap, no “coming soon”.
- Put a price ladder in front of them. Fixed options rather than an open text box, so the answers are comparable to each other.
- Include “I wouldn’t pay for this” as a real option. If that is missing, the result is worthless — you have only surveyed people who were willing to pick a price.
- Confirm the email. An unconfirmed address is a click. A confirmed one is a person who came back.
- Count at a threshold, not an average. Someone who picks $49 would also pay $19, so the honest figure is how many would pay at least X.
Is a stated price the same as revenue?
No, and anyone who tells you otherwise is selling something. People say they will pay and then do not. A stated price is not a forecast.
What it is: a floor you can argue with. It costs a person something to name a number with “I wouldn’t pay for this” sitting right beside it, and plenty of them take that option. That refusal is the part that makes the rest mean anything. A test nobody can fail measures nothing.
How many answers is enough?
Fewer than you would think, because you are not estimating a market — you are checking whether anyone at all will name a price. Twenty considered answers from people in the problem beats two thousand signups from people who clicked a button.
The decision it feeds is binary and worth making early: build it, or find out now that the version in your head is not the version anyone wants to buy.