Updated May 2026
To validate a SaaS idea, validate the problem before the product: confirm the pain is severe and recurring, talk to about ten people who have it, then pre-sell — because someone committing money before the product exists is the only honest signal. Stage your effort 2 hours, then 20, then 200, and let most ideas die early on purpose.
Here's the whole approach, step by step.
These are two different things and people skip the first. Before you ask 'will they buy my solution,' ask 'is this problem painful enough that they'd pay anyone to fix it?' A real, tier-one problem is one people are actively losing time or money to. If the problem isn't severe, no product fixes that.
Stage your effort. Two hours of desk research to see if the problem and any competitors even exist. If it survives that, twenty hours talking to potential customers and testing a landing page. Only once you have strong signals do you spend 200 hours building an MVP. Most ideas should die in the first two stages — that's the framework working, not failing.
Line up about ten short conversations with people who have the problem. Come with ~10 questions aimed at one thing: how painful is this, really? After five calls you'll already sense whether it's a must-fix or a nice-to-have. Don't pitch — listen for the unprompted 'oh god, yes, I hate that.'
Surveys and 'would you use this?' lie; people are polite. The real test is whether someone will commit money or a deposit before the product fully exists. If you talk to ten prospects about pricing and five say yes to actually paying when it's ready, you have something. If everyone loves it but nobody pre-pays, you don't.
Once validated, build the smallest version that solves the core problem — days, not months — and ship it straight to the community the problem came from. Real usage tells you what to build next far better than any plan.
People validate the product— “do you like my idea?” — when they should validate the problem— “is this costing you enough that you'd pay to make it stop?” The first question gets you compliments. The second gets you customers. Compliments don't pay.
A good starting point removes a whole stage of this: instead of guessing at a problem, start from one that already recurs across real threads. That's what Pannly's scored briefs give you — then you run the conversations and the pre-sell from there. Browse SaaS ideas already backed by demand to start from evidence instead of a hunch.
Validate the problem first with a couple of hours of research and a handful of customer conversations, then test demand with a landing page and a pre-sell offer. If people commit money before the product is built, the idea is validated faster and more honestly than any survey could manage.
Not really. People answer surveys to be helpful, not to commit. 'Would you use this?' gets a yes far more often than a credit card does. Use conversations to understand the problem and pre-sales to validate demand — surveys at best shape questions, they don't prove a market.
Around ten focused conversations is enough to feel whether a problem is tier-one. The goal isn't statistical significance; it's pattern recognition. If five or more of ten prospects say they'd pay for a solution when it's ready, that's a strong signal to start building the smallest version.
Pannly does the first stage for you: it surfaces problems that already recur across real Reddit and Hacker News threads and scores them on demand, so you start validation from a problem with evidence rather than a guess. You still run the customer conversations and the pre-sell — but from a much better starting point.
Browsing the scored feed is free. $3 unlocks the full brief — refunded automatically if you ship within 30 days.