AGENT SKILL

customer-journey

Map one persona's path through one JTBD — stages, what they're doing/thinking/feeling, touchpoints, pain; mark evidence vs assumption.

License 许可
MIT
最近更新
2026-10-29
Stars
★ 0
Forks
0

01 前言 Front Matter

name
customer-journey
description
Map one persona's path through one JTBD — stages, what they're doing/thinking/feeling, touchpoints, pain; mark evidence vs assumption.

03 SKILL.md 正文

customer-journey

One persona, one job-to-be-done, 5–7 stages. Multiple personas → multiple journeys; don't merge.

Mode: analytical — judgment inside a fixed frame. Keep the output shape; every verdict cites its evidence; unknowns are marked [TBD] or [VERIFY], never filled in.

When to use

  • A team has a persona doc and wants to surface where in the user's flow they're losing.
  • Funnel analytics show drop-off but the team doesn't know why.

Inputs

  • The persona doc (required).
  • The job-to-be-done in JTBD form ("When <situation>, I want to <motivation>, so I can <outcome>.").
  • Optional: existing funnel analytics.

Procedure

  1. Read Customer-journey.md.
  2. Slug from the JTBD's outcome clause: a workflow's ${workflow.slug} wins; otherwise lowercase it, replace each run of characters outside a-z0-9 with -, trim - at the ends, and cut to ≤ 60 characters at a -. An existing file at the target path is updated in place (show the diff), never duplicated.
  3. Draft docs/research/journeys/<slug>.md:
    • JTBD — verbatim in the doc.
    • Stages — 5–7. User-perspective, not product features. Trigger → Search → Evaluate → Onboard → First success → Habit → Tell others is a good default.
    • Per stage — what they're doing, thinking/feeling, touchpoints, pain.
    • Peak pain moments — the 1–2 highest-drop-off stages. These are the leverage points.
    • Evidence vs assumption — mark every stage ✓ Evidence (quote-supported) or ? Assumption (guess).
    • Improvements — top 3 changes sized by impact, with the metric each would move.

Quality bar

  • Stages are user-perspective ("evaluating options"), not product-perspective ("uses comparison page").
  • Every stage marked Evidence or Assumption. Assumption-heavy journeys are flagged for research before action.
  • Peak pain moments named explicitly.
  • One persona per journey. Multi-persona journeys are merged guesses.

Output

  • File: docs/research/journeys/<slug>.md.
  • Summary listing the 2 peak pain moments + the top improvement.

Anti-patterns to avoid

  • Generic "AAARRR" funnel re-labeled as a journey. The stages must reflect this specific JTBD.
  • Stages full of ? Assumption shipped as if they were research.
  • "Delight" stage added as a vanity step. If there's no evidence of delight, don't add it.
  • Improvements sized by "ease" or "what we can ship next sprint" instead of impact.

04 同仓库更多 Skills