Choose a primary and backup direction

Compare the daily work

Take two real job descriptions per direction. Extract actions, inputs, outputs, collaborators and eligibility. A preference for finance does not distinguish trading from audit; willingness to reconcile data is testable. Mark unstated working conditions unknown and investigate them.

Separate constraints from preferences

Check eligibility, schedule conflicts and unaffordable costs before weighing preferences. A weighted score cannot cancel a hard constraint. Try one work-producing task per direction and compare difficulty, motivation and evidence quality. Preserve preparation shared by the primary and backup.

Define when to reconsider

A choice is revisable. Record evidence, unknowns and a review trigger such as two role conversations or one work sample. Ask AI to find unsupported comparisons and conflicting preferences, then retain your own decision rationale.

Worked example · fictional

From a problem to inspectable work

Fictional case: a third-year student has six hours weekly and a course data report, and is choosing analytics or marketing.

1. Break titles into querying, segmentation and experiment briefs.
2. Reuse the report for an analysis memo, then design a marketing experiment and compare engagement.

Choose analytics provisionally, with marketing analytics as backup. Review the two samples against job requirements in two weeks; language requirements remain unknown.

Failure case and repair

Choose from an industry salary ranking.

The ranking does not establish fit, eligibility or available opportunities.

Add two original job descriptions and one work task; mark unknown conditions.