Description
Write each idea as a falsifiable hypothesis with a metric, list the riskiest assumptions, and test them with the smallest artefact or market experiment before building more.
Validated learning treats product ideas as hypotheses. A brief states what the team wants to learn, the hypothesis ('We believe that [solution] for [user] will result in [outcome]; we will know when we see [measurable result]'), how to run it safely, and which measures would invalidate it or stop it. The riskiest assumptions about users, solution and feasibility are listed and the biggest tested with the smallest artefact, which is often not a prototype: an explainer video, landing page, contract MVP (non-binding letters of intent with a preset signature threshold, such as 500), a proxy test on an existing platform, a concierge service, a design comic in interviews, or a deliberately less-than-minimal release to development partners measured by usage. MVPs test business viability, not functional viability. In practice the label 'MVP' often survives without pre-stated thresholds, which removes the rigour.
Attributes
- Discipline
- Experimentation & Causal Inference
- Loss type
- Vestigial
- Stage
- Strategy, Discovery, Definition
- Data
- Mixed
- Effort
- Weeks
Steps
- Frame the opportunity and write each concept as a falsifiable hypothesis with a measurable result; without a metric there is no testable hypothesis.
- List the assumptions that would force a rethink if false (users and problems, solution and behaviour, technical risk) and pick the few biggest.
- Write a brief: what to learn and why, how to run it, whether it is safe, the threshold that justifies further work, and measures that invalidate it or stop it.
- Choose the smallest test: explainer video, landing page, contract MVP with a preset signature count, proxy test on an adjacent system, concierge service, design comic or a limited release.
- Run it with real target users (interviews, intercepts, feature toggles to growing groups, or in-product metrics) for days to weeks.
- Weigh behaviour over stated opinion; interview selected respondents to understand motivations.
- Compare results with the preset threshold and decide to persevere, pivot or drop; roll facts back into the assumption list and design the next test.
Sources
- Validating Product Ideas (Sharon, 2016)
- Mapping Experiences (Kalbach, 2016)
- User Story Mapping (Patton & Economy, 2014)
- Lean Analytics (Croll & Yoskovitz, 2013)
- Universal Principles of Design (Lidwell et al., 2023)
- Product Management for UX People (Crumlish, 2022)
Also known as
- Build-measure-learn
- Validated learning
- Lean Startup experiment
- Minimum viable product experiment
- MVPe
- MVP test
- Minimum-viable product demand test
- Contract MVP
- Letter of intent test
- Lean feature experiment brief
- Experiment charter
- Proxy test on an adjacent system
- Hypothesis statements and lean market experiments
- Explanatory video test
- Limited product release
- Customer development partners
- Pretotype
- Pretotyping
Categories
- Analytics & Behaviour
- Experimentation & Causal Inference