Idea Validation

What this is for

Help a team learn what must be true before it commits a roadmap, engineering capacity, or a polished launch to an idea.

When this helps

• Teams debate the entire idea instead of testing its riskiest belief.
• Interview recruitment copy accidentally pitches the solution and biases the answer.
• Research, experiments, and fake-door pages exist as separate activities with no shared decision rule.
• A promising click or customer quote is treated as proof without checking the audience, behavior, or contradictory evidence.

How it works

Idea Validation is the orchestration layer across Specky’s existing discovery tools. Select an opportunity, feature, roadmap item, strategy card, document, experiment, or describe a new idea. Specky reads the selected ICP and real Product Graph evidence, then ranks desirability, viability, feasibility, and usability assumptions by downside if wrong. Each assumption keeps evidence for and against, a test method, observable success criteria, and a timebox. The same plan produces neutral recruitment copy, behavioral interview questions, and explicit build, reframe, and stop conditions. From the plan you can create a real Alex research campaign or a linked Experiment Tracker record; a fake-door or waitlist page can then be published from that experiment.

Before you start

• The idea or existing product artifact you want to validate.
• The ICP segment expected to benefit, or an honest note that the audience is still unknown.
• The believed customer problem, expected outcome, known evidence, doubts, and constraints.
• Real interviews, feedback, calls, support signals, analytics observations, or product artifacts in the active workspace when available.

What you should see

• A risk-ranked assumption board covering desirability, viability, feasibility, and usability.
• Evidence for and against each assumption, with real Product Graph references and visible gaps.
• The smallest next test with method, description, success criteria, and timebox.
• A LinkedIn recruitment post, follow-up message, candidate-specific outreach, and behavioral interview guide.
• A real Alex campaign and linked experiment when you choose to create them.
• An explicit current decision: not tested, testing, build with confidence, reframe, or stop.

Try it now

1. Open Market → Idea validation and choose Validate an idea.
2. Select an existing product item or name a new idea. Choose one ICP segment whenever possible.
3. Describe what you believe, what outcome should change, what evidence already exists, and what worries you.
4. Choose Build validation plan. Start with the first critical or high-risk assumption—not the easiest assumption to test.
5. Read evidence for and against. Change the evidence state from Untested to Learning, Supported, or Contradicted only after reviewing the underlying signals.
6. Choose Create experiment on the assumption you want to test. Specky creates a real linked Experiment Tracker record with the proposed method, success criteria, and timebox.
7. Open Recruit & interview. Edit or copy the outreach, then choose Create real Alex campaign to create the shareable research campaign.
8. For fake-door, waitlist, prototype, concierge, or feature-interest tests, open the created experiment and add a Validation Page.
9. Return to Decision after the timebox. Record Build with confidence, Reframe the idea, or Stop this bet using the conditions agreed before the test.

A real example

A team believes growth PMs want an automatic weekly explanation of signup changes. The plan identifies the critical desirability assumption: PMs will trust and act on a cited brief without recreating the investigation. Alex interviews five growth PMs about their last signup investigation, while a concierge experiment sends two manually assembled briefs. The team builds only if at least three target PMs use the brief in a real review and ask to receive it again; it reframes if they value anomaly alerts but not the explanation; it stops if the weekly job is not recurring.

Your next move

Use Alex Research for conversations, Experiment Tracker for the test, Validation Pages for qualified intent, and Feature Marketing only after the promise is ready to explain.