Clarifying questions

Clarifying questions turn vague requests into implementable scope before you choose libraries or schemas. Ask about the user and job-to-be-done, success metrics, data shape, permissions, integrations, non-functional constraints, and what is explicitly out of scope.

| Area | Example questions | |------|-------------------| | User & goal | Who acts, what outcome, how often, what failure costs? | | Data | Source of truth, validation rules, retention, PII? | | Permissions | Roles, tenant boundaries, audit needs? | | Constraints | Timeline, team skills, legacy systems, compliance? | | Success | Metric, analytics event, support ticket reduction? |

					"We need export" → Who exports? What format? Max rows? Sync or async?
Authorized only? Empty result? Error messaging? SLA?
				

On interviews: demonstrate question ordering—user outcome first, implementation last. Show how answers change API shape (sync CSV vs async job with email).

Common pitfalls: jumping to React vs Vue; assuming admin-only; skipping mobile or slow-network cases; no definition of success.

The trade-off is a few minutes of discovery versus days of rework when stakeholders meant something different.

Checklist:

• Restate the request in your own words.
• Ask who, what outcome, and how measured.
• Surface permissions and data boundaries.
• Confirm what is explicitly not in v1.