Guide
How to validate SaaS ideas before you write code
SaaS validation is harder than product validation because you are not selling a thing — you are selling a recurring habit with recurring costs attached. The failure mode is not 'nobody wanted it'; it is 'people wanted it once, and serving them cost more than they paid'.
This guide covers the five checks that matter: recurrence, a concierge MVP, a real price test, unit economics including AI and infrastructure cost, and the signals that mean go.
SaaS validation has one extra question: will they come back?
A one-off product only needs a buyer. Subscription software needs a habit. So validating SaaS means testing recurrence as well as desire: does the problem recur often enough that a monthly charge feels smaller than the pain?
Practical test: ask prospects how often the problem occurred in the last thirty days. Weekly or more supports a subscription. Quarterly usually means a service, a one-time tool, or a usage-based price instead.
Run a concierge version before you build
Deliver the outcome manually for three to five paying customers — spreadsheets, calls, copy-paste, whatever it takes. You learn the real workflow, the actual support load, and which features people never ask for. Most SaaS roadmaps shrink dramatically after a concierge round.
Charge for it. A free concierge pilot teaches you nothing about price, and free users tolerate friction paying ones will not.
Test price with a real page, not a survey
Publish a landing page with tiers, a specific number, and a checkout or booking action. Send it to the segment you identified. Conversion from a targeted audience at a stated price is the only pricing evidence that survives contact with reality.
Run one price, not three variants, until you have enough traffic to distinguish them. Underpowered tests produce confident nonsense.
Do the unit economics before the launch
For each customer, add infrastructure, third-party APIs (AI inference is the current killer), payment fees, and support hours in money. Subtract from price. Then ask how many months of that margin it takes to repay your acquisition cost. If the answer is longer than the average customer stays, the model does not work at any traffic level.
Cap or meter expensive variable costs from day one. A flat price on an uncapped variable cost transfers all the downside to you.
Signals that a SaaS idea is validated
Paying pilot customers at your intended price; a repeatable channel with a measurable cost per signup; usage recurring inside the billing period; a cost per customer well below price; and a specific reason customers cannot easily go back to their old workaround.
Run this on your own idea
Samchu-AI does this work for you. Describe your problem in a paragraph and a nine-expert council returns scored solutions, bottom-up costs, S-curve adoption modelling, a customer profile, a competitor radar and a prioritised action plan — in your own region, currency and language. Your first project is free.
Validate my idea free