Goal nudge fatigue suppression and soft-snooze policy

DEC-0074 - Goal nudge fatigue suppression and soft-snooze policy

Context

Current policy already keeps goal optional and caps nudge frequency to 1/day, but repeated dismissals can still feel noisy for users who intentionally postpone goal setup.

Decision

Keep all existing optional-goal rules and add one lightweight adaptive layer:

• Fatigue trigger:
  • if goal_missing = true and user dismisses/ignores goal nudge >=3 times within rolling 14 days,
  • enter goal_nudge_soft_snooze state.
• Soft-snooze behavior:
  • duration: 14 days,
  • popup-style goal nudge is suppressed,
  • only inline lightweight reminder is allowed at valid touchpoints (home | practice_management | profile).
• Early exit:
  • if user opens goal editor manually, or
  • if user saves/updates goal,
  • then soft-snooze ends immediately.
• Re-entry after snooze:
  • if goal is still missing when snooze ends, normal contextual popup eligibility resumes,
  • still governed by daily cap and protected-context suppression rules.

Decision Value

• Reduces nudge fatigue without adding mandatory steps.
• Keeps UX simple and predictable.
• Preserves personalization opportunity while respecting user autonomy.

Rationale

A single soft-snooze state is enough to reduce annoyance and is easier to scale than multiple nudge frequencies or complex segmentation.

Implications

• Product/UX impact:
  • users who are not ready are not repeatedly interrupted by popup nudges.
  • users still keep a clear path to set goal at any time.
• Data/logic impact:
  • track goal_nudge_dismiss_count_rolling_14d and goal_nudge_soft_snooze_until.
• Operational impact:
  • no additional role/process dependency; policy is deterministic and low-maintenance.

Alternatives considered

• Option A: Keep 1/day cap only, no adaptive suppression.
• Option B: Add multi-tier frequency segmentation by cohort/value score.

Open follow-ups

• No blocker for baseline lock; copy experiments can run later without changing policy contract.