Idea validation
How to Validate a SaaS Idea Before Building It
To validate a SaaS idea before building it, identify the riskiest assumption, talk to people who recently experienced the problem, look for costly workarounds and real buying behaviour, and run the smallest experiment that can change your decision. Enthusiasm is not validation; evidence that survives a specific test is.
What SaaS idea validation is—and is not
Validation is the process of reducing uncertainty about a product bet before committing significant build time. It asks whether a specific group has a meaningful problem, whether the problem is urgent enough to change behaviour, and whether there is a credible path to reach and serve those people.
It is not asking friends whether an idea sounds useful, asking people if they would use a hypothetical feature, or getting an AI system to score the idea in isolation. Those activities may generate questions, but they do not prove demand.
1. Write the riskiest assumption first
Describe the idea as a testable belief. Separate the customer, problem, current alternative, promised outcome, and business model. Then ask which assumption would make the whole idea fail if it were wrong.
2. Run interviews about past behaviour
Talk to people who recently faced the problem, not only people who match a demographic profile. Ask what happened last time, what they tried, what it cost, who was involved, and what they did next. Keep the idea out of the conversation until you understand the current behaviour.
- Tell me about the last time this happened.
- What did you try before looking for something else?
- What was the most frustrating or expensive part?
- How are you solving it today?
- Who would need to approve a change or purchase?
- What would make you switch from the current workaround?
3. Separate interest from commitment
A positive interview is a lead, not a conclusion. Stronger signals include a customer sharing a real workflow, introducing a decision-maker, committing to a pilot, giving access to relevant data, agreeing to a follow-up, placing a deposit, or paying for a constrained version of the solution.
The right signal depends on the business. The important part is that the person gives up something scarce—time, money, access, reputation, or an existing workaround—in response to the problem you are testing.
4. Run the smallest honest experiment
Choose a test that exposes the riskiest assumption without pretending the product already exists. Depending on the idea, that could be a concierge service, a waitlist with a clear commitment, a fake-door test, a prototype task, a pre-sale conversation, or a manual workflow delivered to a small group.
Define the pass, pivot, and stop conditions before you run the test. A result is useful when it can change what you do next, including deciding not to build the original version.
- 01
Pass
The evidence supports the assumption and justifies a narrower build or paid pilot.
- 02
Reframe
The problem is real but the segment, job, or proposed solution needs to change.
- 03
Stop
The evidence is too weak or the problem is not costly enough to justify continuing.
5. Record the decision and the evidence
Write down what you believed, what you tested, what happened, what remains unknown, and what you will do next. This turns validation into a compounding asset rather than a set of conversations that disappear into notes and memory.
If the evidence says “not yet,” that is a useful result. It protects the team from spending months polishing a solution before it has earned the right to exist.
Where Specky fits
Keep the evidence attached to the work.
Specky connects customer signals, research, decisions, specs, tickets, and outcomes in one Product Graph. It drafts and connects the work; your team reviews and decides.
Questions people ask about this topic
How many interviews do I need to validate a SaaS idea?+
There is no magic number. Interview enough people in the target situation to see whether the problem, workaround, and language repeat, then run an observable test of the riskiest assumption. The quality and relevance of the evidence matter more than a fixed quota.
Can AI validate my startup idea for me?+
AI can help research a market, identify assumptions, draft neutral interview questions, and synthesize evidence. It cannot replace conversations with real people or create willingness to pay. Treat AI output as a hypothesis and verify it with real behaviour.
What counts as real validation?+
Real validation is evidence that changes confidence in a specific assumption: repeated costly problems, existing workarounds, a credible buyer, a paid pilot, a deposit, or another action that carries a real commitment.
What is Vouch?+
Vouch is the idea-validation workspace from the Specky team. It helps founders structure assumptions, get AI coaching, and run real interviews before they commit to a build. Visit vouch.specky.space to explore it.
Continue exploring
Vouch
Structure assumptions and run real interviews before you build.
Specky idea validation
Connect validation assumptions, research campaigns, and experiments.
JTBD interview template
Use 21 practical Jobs-to-be-Done interview questions.
For solo founders
Run a grounded product workflow before hiring a PM team.