WorthBuilding guide

How to validate a SaaS idea without building the whole product

SaaS idea validation is a sequence of questions: does the problem happen, can you improve the workflow, can you reach the user, and is there a buyer? A prototype can answer some of these. Compliments and vote totals cannot answer all of them.

Make the assumptions testable

Describe the audience, trigger, current workaround, and proposed outcome. Then write the assumptions that would make the idea fail if they were wrong. For a tutorial maintenance tool, these might be that teams have enough tutorials to make updates painful, know when screens change, and would let a tool access the relevant product environment.

Pick the least certain assumption first. Testing a polished interface is not useful if you have not established whether you can access the required data or reach the person responsible for the workflow.

Interview people with a recent example

Ask people to describe the last time the problem occurred. Compare your notes across conversations: do they use similar workarounds, experience similar consequences, and have a reason to change? Record disagreement as carefully as enthusiasm.

Avoid leading questions such as ‘Would you pay for an app that saves you hours?’ Ask what they currently spend, who approves tools, and how a previous purchase was decided. Their actual process can reveal constraints that a hypothetical price question misses.

Run a small manual pilot

Deliver the proposed output with existing tools before automating every step. For a support-to-help-article idea, draft an article from a resolved question and let the support team review it. For site-visit quotes, produce a draft from a sample the customer is allowed to share.

Agree on the scope, turnaround, permission to use the input, and what the customer will review. Never represent a manual service as an autonomous product. The point is to learn where the output helps and where it breaks.

  • Did the person supply the input needed to try it?
  • Did they use the result in their real workflow?
  • What did they have to correct?
  • Did they return when the task occurred again?

Test commitment with a clear offer

Once the outcome is useful, present a specific offer with a defined scope and price. A paid pilot can provide stronger evidence than a waitlist, but only promise work you can deliver and use clear terms. A free pilot can still teach you about use; label it accurately in your notes.

Separate interest, trial participation, repeat use, and payment. An ‘I'd pay’ click on WorthBuilding is a conversation lead, not a purchase or verified budget. Upvotes may also include AI evaluation points, so they must not be presented as a count of prospective customers.

Choose a next step from the evidence

Build when you have a narrow repeatable workflow and people willing to use its result. Reframe when the problem matters but the proposed solution does not fit. Pause when you cannot reach the audience, the input is unavailable, or existing tools already solve the job well.

Keep an experiment log with the date, audience, test, observation, and next decision. It is easier to notice weak evidence when it is written down separately from your hopes for the product.

Explore problems to investigate

Keep exploring