Product Experiment
A product experiment is a deliberate test that exposes a product idea or assumption to real users or market conditions to learn whether it holds before the team commits significant time or money. Each experiment starts from a stated hypothesis, uses a defined method and measure, and ends with evidence that informs a decision: proceed, change course or stop.
How Product Experiments Work
Every product experiment follows the same loop. The team picks a risky assumption, states it as a testable hypothesis, runs the smallest test that can produce useful evidence, then reviews the result and decides what to do next. Eric Ries described this cycle in The Lean Startup as build, measure, learn.
Experiments differ mainly in what they test and how strong the resulting evidence is. Strategyzer's Testing Business Ideas groups the risks into three kinds:
- Desirability. Do customers want this?
- Feasibility. Can we build and deliver it?
- Viability. Can it sustain a business at a price customers will pay?
Methods range from cheap and fast to slow and rigorous. Early, discovery-oriented experiments include problem interviews, landing pages that measure sign-ups for a product that does not exist yet, and "concierge" tests where the team delivers the service by hand. Later, validation-oriented experiments include prototype tests, paid pilots and A/B tests on a live product. Ideas that are riskier and cheaper to test usually go first, which is why teams often use assumption mapping to choose what to test.
Why Product Experiments Matter
Teams consistently overrate their own ideas. Experiments make it cheap to be wrong: a week spent testing demand is far less costly than a quarter spent building a feature nobody adopts. They also change the conversation inside a team, from whose opinion wins to what the evidence shows, and they leave a record of why a decision was made.
An experiment is only useful if its result can change a decision. A test whose outcome would not alter the plan, or whose result can be read either way, is activity rather than learning.
For a view of how experiments fit into the full path from idea to backlog, see validating a SaaS idea before you build.
Product Experiment Example
A team behind a B2B scheduling tool is considering a paid "team analytics" add-on. Before building it, they add an "Analytics" item to the settings menu for a month. Clicking it shows a short description and a "Join the pilot" button. They track how many admins click and how many join. Enough admins join that the team runs a second experiment: for six pilot accounts, an analyst produces the reports by hand each week, and the team checks whether admins open and act on them. Only after both tests show real use does the add-on enter the roadmap.
Product Experiment vs. Experiment Design
The two terms describe different levels. A product experiment is the learning activity itself: the test the team runs and the decision it informs. Experiment design is the plan that structures that test, covering the hypothesis, method, participants, metric, duration and success criteria set in advance. Good experiment design is what makes a product experiment's result trustworthy.