3) Create recruiting plan + screener + outreach copy
3) 制定招募计划 + 筛选问卷 + outreach文案
Inputs: Participant profile; available channels (CRM, support, community, ads, LinkedIn).
Actions: Draft outreach messages, a screener, and scheduling logistics. Expect high drop-off; plan volume accordingly.
Outputs: Recruiting plan + screener + outreach copy.
Checks: Screener screens for the story you need (recency, context, alternatives), not “interest in our product.”
输入信息：参与者画像；可用渠道（CRM、客服、社区、广告、LinkedIn）
行动：撰写招募信息、筛选问卷和日程安排。预计会有较高的流失率，需相应规划招募量。
输出成果：招募计划 + 筛选问卷 + outreach文案
检查要点：筛选问卷针对您需要的故事（时效性、背景、替代方案），而非“对我们产品的兴趣”
4) Draft the interview guide (story-first)
4) 撰写访谈指南（以故事为核心）
Inputs: Interview type + hypotheses/unknowns.
Actions: Build a guide that elicits specific stories (“last time…”) and avoids leading questions. Include probes, pivots, and time boxes. Add consent/recording script.
5) Run interviews + capture clean notes (PM/Design present)
5) 开展访谈 + 记录清晰笔记（产品经理/设计师到场）
Inputs: Guide, logistics, notes template.
Actions: Run the session, follow the story, and capture verbatims. If possible, have PM + design observe live (or listen to recordings) to avoid secondhand dilution.
7) Synthesize across interviews into themes and opportunities
7) 整合多场访谈内容，提炼主题与机会点
Inputs: Tagged notes across interviews.
Actions: Cluster by outcomes/struggles; capture contradictions; quantify lightly (counts) without over-claiming. Translate insights into opportunities and recommendations with confidence levels.
Outputs: Synthesis report + opportunity list.
Checks: Every major insight has at least 2 supporting interviews (or is labeled “single anecdote”).
8) Share, decide, follow up, and run the quality gate
8) 分享成果、做出决策、跟进并进行质量校验
Inputs: Draft pack.
Actions: Produce a shareable readout, propose next steps, and create a lightweight customer panel habit (5–10 engaged users). Run references/CHECKLISTS.md and score with references/RUBRIC.md.
Outputs: Final User Interview Pack + Risks/Open questions/Next steps.
Checks: Stakeholders can restate (a) key learning, (b) decision implication, (c) what happens next.
Example 1 (Discovery): “I’m redesigning onboarding for a B2B product. Create a recruiting plan + discussion guide for 8 discovery interviews with new trial users.”
Expected: participant criteria, outreach + screener, discovery guide, notes template, synthesis plan, and a ready-to-run pack.
Example 2 (Switch/JTBD): “We lose deals to spreadsheets. Run switch interviews to learn what triggers teams to move off spreadsheets and what they try instead.”
Expected: switch interview guide (timeline + forces), recruiting criteria emphasizing recency, and a synthesis structure that outputs ‘push/pull/anxieties/habits’.
Boundary example: “Ask users what features they want and build whatever they say.”
Response: redirect to story-based interviewing; clarify decision context; avoid feature-request interviews without behavioral grounding.