Product discovery is the set of activities that determine what to build before committing engineering resources to build it. It is the antidote to the most expensive mistake in product development: building the wrong thing well. A team of ten engineers working for six months on a product that users do not want has wasted five person-years of highly skilled labour. Four weeks of structured discovery that reveals the mismatch before development begins is the highest-ROI investment in the product development lifecycle.
Discovery operates in three risk categories. Value risk: will users want this? Usability risk: will they be able to use it? Feasibility risk: can we build it? Business viability risk: should we build this? Each requires different discovery techniques.
For value risk, there is no substitute for talking to potential users early and often. Not showing them what you plan to build and asking "would you use this?" — the answer is almost always yes, because it is socially awkward to say no. Instead, understanding their current behaviour: what do they do today, what frustrates them, what solutions have they tried and abandoned. The target customer who has already tried multiple inadequate solutions to a problem is your best early adopter.
Prototyping is the primary technique for testing usability before building. A clickable Figma prototype that simulates the key user journeys can be tested with five to ten potential users in a few hours, revealing where the proposed design confuses or frustrates users before a line of code is written. The rule of thumb is that five users reveal eighty percent of usability issues.
Feasibility spikes — time-boxed technical investigations where engineers explore whether a specific technical approach is viable — answer feasibility risk without committing to full implementation. A two-day spike that reveals a critical third-party API has unacceptable latency prevents six weeks of architecture built on a false assumption.