Back to Details

behavioral-product-design

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

Behavioral Product Design

Behavioral Product Design

Scope

适用范围

Covers
  • Turning a desired user behavior into an executable design + experiment plan
  • Diagnosing behavior using barriers/drivers (motivation, ability/friction, uncertainty, habit, context)
  • Designing behavioral interventions (e.g., defaults, commitment devices, loss aversion/progress, reducing uncertainty) with ethical guardrails
  • Producing decision-ready artifacts a PM/Design/Eng team can build and test
When to use
  • “Help me apply behavioral science / behavioral economics to this flow.”
  • “We need to improve retention / activation / onboarding completion.”
  • “Design a streak / habit loop / reminder system (without being spammy).”
  • “Users procrastinate (present bias). How do we get them to do the thing?”
  • “People stick with the status quo. How do we drive switching/adoption?”
  • “Users are uncertain / anxious. How do we reduce uncertainty and move them forward?”
When NOT to use
  • You need upstream strategy first (vision, positioning, roadmap). Use 
    defining-product-vision
    / 
    prioritizing-roadmap
    .
  • You can’t name the target user + target behavior + success metric (this becomes generic advice).
  • The goal is to create dark patterns (deception, coercion, addiction, hidden costs). Don’t do this.
  • The domain is regulated/high-stakes (medical, financial advice, minors). Require domain/legal review and tighter safeguards.
涵盖内容
  • 将期望的用户行为转化为可执行的设计+实验方案
  • 利用阻碍/驱动因素（动机、能力/摩擦、不确定性、习惯、场景）进行行为诊断
  • 设计带有伦理约束的行为干预方案（如默认设置、承诺机制、损失厌恶/进度展示、降低不确定性）
  • 产出可供产品经理/设计/开发团队直接落地测试的决策就绪型文档
适用场景
  • “帮我将行为科学/行为经济学应用到这个流程中。”
  • “我们需要提升用户留存/激活/新用户引导完成率。”
  • “设计一个打卡/习惯养成循环/提醒系统（避免垃圾信息）。”
  • “用户存在拖延行为（即时偏差）。如何让他们完成目标动作？”
  • “用户倾向于维持现状。如何推动他们切换/采用新功能？”
  • “用户存在不确定/焦虑情绪。如何降低不确定性并促进行动？”
不适用场景
  • 你首先需要上游战略（愿景、定位、路线图）。请使用
    defining-product-vision
    / 
    prioritizing-roadmap
    。
  • 你无法明确目标用户+目标行为+成功指标（此时给出的建议会过于通用）。
  • 目标是创建“暗黑模式”（欺骗、强制、成瘾、隐藏成本）。请勿这样做。
  • 领域受监管/高风险（医疗、财务建议、未成年人相关）。需要领域/法务审核及更严格的保障措施。

Inputs

输入要求

Minimum required
  • Product context + target user segment
  • The target behavior (what user action you want more of, in what context)
  • Baseline funnel/retention metrics (even rough) + where the drop happens
  • Constraints: platform (web/mobile), notification channels, brand/tone, time box
  • Existing evidence: user research notes, support tickets, analytics, session replays (if any)
Missing-info strategy
  • Ask up to 5 questions from references/INTAKE.md.
  • If answers aren’t available, proceed with explicit assumptions and label unknowns. Offer 2 scopes: narrow (1 behavior) vs broad (journey).
最低必要输入
  • 产品背景+目标用户群体
  • 目标行为（你希望用户在特定场景下更多完成的动作）
  • 基准转化漏斗/留存指标（即使是粗略数据）+ 流失节点
  • 约束条件：平台（网页/移动端）、通知渠道、品牌调性、时间限制
  • 现有证据：用户研究笔记、支持工单、分析数据、会话回放（如有）
缺失信息处理策略
  • 从references/INTAKE.md中最多提出5个问题。
  • 如果无法获取答案，基于明确假设推进并标注未知信息。提供两种范围选项：窄范围（单个行为） vs 宽范围（全旅程）。

Outputs (deliverables)

输出成果（交付物）

Produce a Behavioral Product Design Pack (in-chat as Markdown; or as files if requested), in this order:
  1. Context snapshot (goal, segment, constraints, baseline)
  2. Target behavior spec (behavior statement + success metric + guardrails)
  3. Behavioral diagnosis (barriers/drivers; where bias/friction/uncertainty shows up)
  4. Intervention map (ideas mapped to journey moments + mechanism + risk)
  5. Prioritized intervention shortlist (top 1–3 with rationale)
  6. Behavioral design specs (1–3 build-ready “intervention cards”)
  7. Experiment + instrumentation plan (events, primary/guardrail metrics, rollout/rollback)
  8. Risks / Open questions / Next steps (always included)
Templates: references/TEMPLATES.md
产出Behavioral Product Design Pack（默认在对话中以Markdown格式呈现；如需可生成文件），顺序如下：
  1. 背景快照（目标、用户群体、约束条件、基准数据）
  2. 目标行为规范（行为说明+成功指标+约束规则）
  3. 行为诊断（阻碍/驱动因素；偏差/摩擦/不确定性出现的节点）
  4. 干预地图（对应旅程节点的想法+机制+风险）
  5. 优先级干预短名单（排名前1-3的方案及理由）
  6. 行为设计规范（1-3个可直接落地的“干预卡片”）
  7. 实验与埋点方案（事件、核心/约束指标、灰度发布/回滚策略）
  8. 风险/未解决问题/下一步行动（必须包含）
模板：references/TEMPLATES.md

Workflow (8 steps)

工作流程（8步骤）

1) Intake + define the target behavior

1) 需求收集+定义目标行为

  • Inputs: User context; references/INTAKE.md.
  • Actions: Clarify the user, context, and one primary target behavior. Define success + guardrails (what must not get worse).
  • Outputs: Context snapshot + target behavior spec.
  • Checks: Target behavior is observable and time-bounded (not “be more engaged”).
  • 输入：用户提供的背景；references/INTAKE.md
  • 动作：明确用户、场景及一个核心目标行为。定义成功指标+约束规则（哪些指标不能恶化）。
  • 输出：背景快照+目标行为规范
  • 检查项：目标行为可观察且有时间边界（而非“提升用户参与度”这类模糊表述）

2) Map the current journey + “moments that matter”

2) 绘制当前用户旅程+“关键节点”

  • Inputs: Current flow/JTBD; baseline funnel.
  • Actions: Sketch the steps from trigger → action → outcome. Mark drop-offs and emotional moments (uncertainty, effort, waiting, completion).
  • Outputs: Journey map summary + top 3 friction points.
  • Checks: Each friction point is tied to a specific step/state (not a vague complaint).
  • 输入：当前流程/用户任务（JTBD）；基准转化漏斗
  • 动作：勾勒从触发→动作→结果的步骤。标记流失节点及情绪节点（不确定性、费力、等待、完成）。
  • 输出：用户旅程地图摘要+Top3摩擦点
  • 检查项：每个摩擦点都对应具体步骤/状态（而非模糊抱怨）

3) Run a behavioral diagnosis (barriers + drivers)

3) 开展行为诊断（阻碍+驱动因素）

  • Inputs: Journey moments; evidence; assumptions.
  • Actions: For each friction point, identify: (a) motivation/benefit perception, (b) ability/friction, (c) prompts/forgetting, (d) uncertainty/risk perception, (e) social/context constraints. Map likely mechanisms (e.g., present bias, status quo, uncertainty aversion, loss aversion/progress).
  • Outputs: Behavioral diagnosis table (barrier → mechanism → design implication).
  • Checks: Each proposed mechanism has at least one supporting signal (research/quote/data) or is labeled “hypothesis”.
  • 输入：用户旅程节点；现有证据；假设
  • 动作：针对每个摩擦点，识别：(a) 动机/收益感知，(b) 能力/摩擦，(c) 触发/遗忘，(d) 不确定性/风险感知，(e) 社交/场景约束。匹配可能的机制（如即时偏差、现状偏差、不确定性厌恶、损失厌恶/进度展示）。
  • 输出：行为诊断表格（阻碍→机制→设计启示）
  • 检查项：每个提出的机制至少有一个支持信号（研究/引用/数据），或标注为“假设”

4) Generate intervention ideas (mechanism-first, not UI-first)

4) 生成干预方案（基于机制而非UI）

  • Inputs: Diagnosis table.
  • Actions: Brainstorm 2–4 interventions per priority barrier using the pattern library in references/WORKFLOW.md (defaults, reducing uncertainty, progress/loss framing, commitment devices, reminders, celebration/pause moments).
  • Outputs: Intervention inventory (10–20 ideas) with mechanism tags.
  • Checks: At least one idea reduces friction (ability) and one reduces uncertainty (trust), not only “add reminders”.
  • 输入：行为诊断表格
  • 动作：针对每个优先级阻碍，利用references/WORKFLOW.md中的模式库头脑风暴2-4个干预方案（默认设置、降低不确定性、进度/损失框架、承诺机制、提醒、庆祝/暂停节点）。
  • 输出：干预方案清单（10-20个想法）及机制标签
  • 检查项：至少有一个方案减少摩擦（能力），一个方案降低不确定性（信任），而非仅“添加提醒”

5) Add resilience + reinforcement (without manipulation)

5) 增强韧性+强化（避免操纵）

  • Inputs: Intervention inventory.
  • Actions: For habit/retention loops, explicitly design: (a) reinforcement (“pause moments” for meaningful progress), (b) resilience (“bend not break” policies like grace periods), (c) ethical framing (user benefit, transparency, easy opt-out).
  • Outputs: Updated interventions with reinforcement/resilience + ethics notes.
  • Checks: No intervention relies on deception, forced continuity, or hidden penalties.
  • 输入：干预方案清单
  • 动作：针对习惯/留存循环，明确设计：(a) 强化（用于展示有意义进度的“暂停节点”），(b) 韧性（如宽限期这类“灵活而非断裂”的规则），(c) 伦理框架（用户收益、透明度、轻松退订）。
  • 输出：更新后的干预方案，包含强化/韧性+伦理说明
  • 检查项：无干预方案依赖欺骗、强制连续性或隐藏惩罚

6) Prioritize and pick the top 1–3 bets

6) 优先级排序并选出Top1-3方案

  • Inputs: Updated inventory; constraints.
  • Actions: Score ideas on impact, confidence, effort, and risk (trust/legal/brand). Pick 1–3 that cover different failure modes (friction vs uncertainty vs motivation).
  • Outputs: Prioritized shortlist + “why these” rationale.
  • Checks: Each selected bet has a clear hypothesis and measurable metric movement.
  • 输入：更新后的方案清单；约束条件
  • 动作：从影响、信心、实施成本、风险（信任/法务/品牌）维度对方案打分。选出1-3个覆盖不同失败模式（摩擦vs不确定性vs动机）的方案。
  • 输出：优先级短名单+“选择理由”
  • 检查项：每个选中的方案都有明确假设及可衡量的指标变化

7) Write build-ready behavioral design specs + experiment plan

7) 撰写可落地的行为设计规范+实验方案

  • Inputs: Shortlist; references/TEMPLATES.md.
  • Actions: For each bet, write an intervention spec: hypothesis, mechanism, UX/copy, states, edge cases, instrumentation, rollout/rollback, and guardrails.
  • Outputs: 1–3 behavioral design specs + experiment/instrumentation plan.
  • Checks: Engineering can implement without major missing decisions; measurement is feasible.
  • 输入：优先级短名单；references/TEMPLATES.md
  • 动作：针对每个方案，撰写干预规范：假设、机制、UX/文案、状态、边缘案例、埋点、灰度发布/回滚、约束规则。
  • 输出：1-3个行为设计规范+实验与埋点方案
  • 检查项：开发团队无需补充大量决策即可实施；可进行有效测量

8) Quality gate + finalize

8) 质量审核+最终定稿

  • Inputs: Draft pack.
  • Actions: Run references/CHECKLISTS.md, score with references/RUBRIC.md, and add Risks / Open questions / Next steps.
  • Outputs: Final Behavioral Product Design Pack.
  • Checks: The pack is specific to this product and can be executed in 1–2 sprints.
  • 输入：初稿包
  • 动作：使用references/CHECKLISTS.md和references/RUBRIC.md进行审核打分，并添加风险、未解决问题、下一步行动。
  • 输出：最终Behavioral Product Design Pack
  • 检查项：该包针对当前产品定制，可在1-2个迭代周期内落地

Quality gate (required)

质量审核（必填）

  • Use references/CHECKLISTS.md and references/RUBRIC.md.
  • Always include: Risks, Open questions, Next steps.
  • 使用references/CHECKLISTS.md和references/RUBRIC.md。
  • 必须包含：风险、未解决问题、下一步行动。

Examples

示例

Example 1 (Activation): “New users abandon setup on step 3. Use behavioral science to redesign onboarding and propose 2 experiments.”
Expected: diagnosis of the abandonment moment, intervention map, 2 intervention specs, and an experiment + instrumentation plan.
Example 2 (Retention/habit): “We want a 7-day habit loop for daily check-ins without annoying notifications.”
Expected: habit/reinforcement plan (incl. bend-not-break), celebration moments, a streak spec, and guardrail metrics.
Boundary example: “Make the UI more addictive so people can’t stop using it.”
Response: refuse dark patterns; reframe toward user-beneficial behaviors, transparency, and opt-out controls.
示例1（激活）：“新用户在步骤3放弃设置。使用行为科学重新设计新用户引导并提出2个实验方案。”
预期产出：流失节点诊断、干预地图、2个干预规范、实验与埋点方案。
示例2（留存/习惯）：“我们希望打造7天习惯养成循环，用于每日签到，且不发送烦人的通知。”
预期产出：习惯/强化计划（包含灵活而非断裂的规则）、庆祝节点、打卡规范、约束指标。
边界示例：“让UI更具成瘾性，让用户无法停止使用。”
回应：拒绝暗黑模式；重新聚焦于用户受益的行为、透明度及退订控制。