The most expensive way to validate a SaaS idea is to build it. The second most expensive is to ask friends what they think. Here is what actually works.
The only validation signal that matters
Someone hands you money — or a credible commitment of money — for something that doesn't fully exist yet. Everything else is a proxy for this, and most proxies are unreliable.
A five-step validation playbook
- Define the exact person and the exact problem in one sentence each.
- Write a landing page that promises a specific outcome for that person.
- Have ten conversations with that person, ideally on video.
- Run a small, targeted ad or outreach campaign to that landing page.
- Try to take pre-orders, deposits, or letters of intent.
What good validation looks like
- You can describe the customer without hand-waving.
- They already spend money solving this problem, badly.
- They can articulate what a solution is worth to them.
- Multiple people have said "how do I buy this" unprompted.
What weak validation looks like
- "That's a cool idea" from friends.
- Newsletter signups without follow-through.
- Interest in the abstract with no willingness to commit.
When to start building
When you have credible signals that at least a handful of people will pay for the outcome you're describing. Not when you're excited. Not when the design is done. When someone else's money is on the table or realistically imminent.
Frequently asked questions
How long should validation take?+
For most B2B SaaS ideas, two to four weeks of focused effort is enough to know whether to keep going.
Can you validate without a landing page?+
You can validate faster with one. Ten hours of setup often saves ten weeks of building the wrong thing.
What if no one wants it?+
Then you've saved yourself six months and $30k. That's a win, not a failure.