An idea becomes a project only when the question is clear.

The decisive early step is not selecting a solution, but translating a vague topic into a problem space with affected people, a current state, hypotheses, boundaries and a testable working question.

From an open topic to a workable question

A problem space must be open and bounded at the same time: open enough for alternatives to remain visible, narrow enough for research to stay purposeful.

01 · OpenDetach the topic from the first solution and surface alternative causes, perspectives and affected people.

02 · BoundMake scope, non-scope, constraints and decision-relevant knowledge gaps explicit.

03 · CondenseTurn knowledge, assumptions and hypotheses into a solution-open, testable project question.

CORE METHOD

The problem space

Initial question → people affected → current state → desired change → assumptions/hypotheses → constraints → scope/non-scope → project question. Targeted research starts only then.

THE PRACTICE TEST

Write your project idea without product, tool or model names. If you can then describe the problem, affected people, desired change and boundaries, you have a working space rather than merely a solution wish.

NEXT ARTICLE

Article 3 · Deep Research is a chain, not a button
Multi-stage research with AI.