THE STARTUP JOURNEY / 02 / Build
What should we build?
Build only what is necessary to answer the next important question.
Evidence that counts
Observed workflow needs, prototype task completion, feasibility evidence, and a bounded product hypothesis.
Signals to question
A long feature list, polished screens, or a technically impressive architecture.
The principal risk
Spending development time on assumptions that a simpler experiment could resolve.
Your decision gate
Does the smallest proposed solution address a demonstrated workflow and test the riskiest assumption?
The next proof
Test a focused prototype with the intended user. Separate usability problems from lack of demand.
Set the success criterion
Define the task, expected behaviour, and acceptable failure conditions before testing.
Do not do yet
Feature expansion, premature infrastructure, and solving several customer problems at once.
Choose the decision the evidence supports
Proceed when the gate is met. Iterate when the approach needs changing. Stop the current path when evidence undermines it. If observations are missing or conflicting, record insufficient evidence and define the next test.
Connect the decision to capital
Account for the cost of the next experiment, the time it consumes, and the commitments it creates. Product, commercial, and ownership decisions should be assessed together. Use specialist review where technical, financial, or legal judgement is required.