仪表板描述生成助手

仪表板名称：用户增长与稳定性监控

一、仪表板目的

• 全局监控核心健康度：规模（DAU、新增）、粘性（D1/D7 留存、会话时长）、稳定性（崩溃率）。
• 快速定位异常，支撑增长、运营、研发对不同环节的高效决策。
• 提供按渠道、版本、平台等维度的对比与拆解，明确问题来源与改进优先级。

二、关键问题

• 规模：DAU是否健康增长？增长来自新增还是回访？
• 质量：新增用户能否在D1/D7留存？会话时长是否稳定？
• 稳定性：崩溃率是否上升？是否集中在特定版本或机型？
• 归因：异常发生在何时、哪个渠道/版本/地区/平台？影响范围多大？

三、指标口径（保持一致且在仪表板内可查看定义）

• DAU：当日至少一次活跃的去重用户数。
• 新增用户：首次完成激活的用户数（激活口径建议在系统内可配置，如首次打开/注册/绑定设备；维持单一口径）。
• 留存率D1/D7：某日新增用户在其+1/+7天仍有活跃的比例。计算：第t天回访人数 ÷ 该日新增用户数（按新增当日分群）。
• 会话时长：单次会话持续时间的中位数与均值；会话定义为用户连续操作，间隔超过30分钟视为新会话（阈值可配置）。
• 崩溃率：崩溃会话数 ÷ 总会话数；同时提供“崩溃用户率”（发生过崩溃的活跃用户 ÷ DAU）作为参考。两者需标注明确且不可混用。

四、可视化与布局

• 顶部KPI卡（当日/昨日/环比/7日均）：DAU、新增、D1/D7留存、会话时长（均值/中位数）、崩溃率。
• 趋势区：7/14/28天趋势，对DAU、新增、留存率、会话时长、崩溃率分别绘制，支持周内季节性标记与节假日标注。
• 留存分群热力图：按“新增日期分群”展示D1、D7留存走势，支持按渠道/版本过滤。
• 质量与稳定性深挖：
  • 会话时长分布（分位数、箱线图），按平台/渠道对比。
  • 崩溃率趋势与Top版本/机型/错误类型榜单，支持下钻到具体版本。
• 归因维度筛选：日期、平台（iOS/Android/小程序等）、应用版本、渠道来源、地区/国家、用户类型（新/回访）。

五、解读与行动指南

• DAU与新增：若DAU增长主要来自新增但留存下降，优先优化新手引导与首日关键路径；若回访下滑，检查推送/内容更新节奏与功能可用性。
• 留存D1/D7：D1反映首日体验与价值感知，D7反映产品习惯建立。若D1正常、D7下滑，补强长期价值与召回触点；若D1低，检查首登流程、性能与崩溃是否影响激活体验。
• 会话时长：均值受极端值影响，重点关注中位数与P75。异常波动需与版本发布、渠道变化同步核对。
• 崩溃率：先看整体趋势，再定位高占比版本/机型/错误。评估受影响会话/用户规模与留存/时长的关联，指导修复优先级。
• 组合判断：同时监控“新增↑且崩溃率↑且D1↓”，提示质量问题正在稀释增长。

六、阈值与预警（建议）

• 阈值设定：以近4周同周期均值为基线，结合±2个标准差触发预警；另提供业务阈值（可配置）。
• 预警类型：
  • 留存率：D1或D7相较基线下降超阈值。
  • 崩溃率：日内连续N个小时上升超阈值，或特定版本崩溃率显著高于总体。
  • 会话时长：中位数骤降并伴随崩溃率上升。
• 通知方式：邮件/IM推送，带上影响范围（用户数、会话数）与Top归因维度。

七、数据与治理

• 刷新频率：建议小时级快照+日终固化；留存数据随时间回补，以T+1为准。
• 口径管理：在仪表板内提供“指标字典”，记录计算方式、窗口定义、事件名、去重逻辑与版本变更说明。
• 版本兼容：发布新版本时标记里程碑，确保对比分析准确。
• 隐私与权限：按角色控制渠道、地区与明细的访问范围。

八、落地使用场景

• 日常例会：5分钟看顶部KPI与异常提示，10分钟定位渠道/版本，明确当日动作（修复/召回/投放调整）。
• 发布回溯：版本发布后持续跟踪72小时崩溃率、D1与会话中位数，评估是否需要灰度扩大或紧急回滚。
• 渠道评估：按渠道看新增质量（D1/D7、时长），优化投放与预算分配。

结论 该仪表板以“规模-粘性-稳定性”三环为核心，通过标准化口径、分群留存、版本与渠道下钻以及异常预警，帮助团队快速判断产品健康度、定位问题成因、确定优先动作，从而在可控质量下实现有效增长。

Dashboard description: GMV, CAC, LTV, Repurchase Rate, Inventory Turnover Days, Gross Margin

Purpose and audience

• Purpose: Provide a single, trusted view of growth efficiency, customer value, and operational health to support decisions on marketing spend, pricing, inventory allocation, and profitability.
• Primary users: Executive team, Finance, Growth/Marketing, Merchandising, Supply Chain.

Top-level view (default)

• KPI tiles with period value, trend (WoW/MoM), and YoY deltas: GMV, CAC, LTV, Repurchase Rate, Inventory Turnover Days (DIO), Gross Margin.
• Contextual ratios and flags under KPI tiles: LTV:CAC ratio and CAC payback (months), GMV vs Target, DIO vs Target, Margin vs Target. Color-coded thresholds are configurable.
• Time series: GMV and Gross Margin (%) over time; CAC and LTV over time; DIO over time.
• Cohort panels: LTV by acquisition cohort; Repurchase Rate by first-purchase cohort (e.g., 30/60/90-day windows).
• Drivers/Breakdowns: By channel, campaign, product category, region, and customer segment (new vs returning).

Metric definitions and calculations Note: Metrics align to the finance calendar and accounting policies; filters allow selection of order date vs revenue recognition date. Defaults are below.

• GMV (Gross Merchandise Value) Definition: Total value of customer orders placed in the selected period, before taxes and shipping, after discounts, excluding cancellations and returned orders. Formula: Sum(order_line_price_after_discounts × quantity) for completed orders.

• CAC (Customer Acquisition Cost) Definition: Fully loaded acquisition spend divided by the number of first-time purchasers acquired in the period. Formula: (Marketing + Sales acquisition spend attributable to new customers) / count(new customers with first purchase date in period). Notes: Attribution window and spend components are documented; defaults to cash basis spend in the same period as first purchase.

• LTV (Customer Lifetime Value) Definition (default): Realized 12-month gross profit per customer by acquisition cohort. Formula: For each cohort, LTV_12M = sum over months 1–12 of (Net Revenue × Gross Margin%) per customer, divided by customers in cohort. Optional discount rate can be applied. Notes: Predictive LTV can be enabled; reported separately from realized LTV.

• Repurchase Rate Definition: Share of new customers who make at least one additional purchase within the selected window (default: 90 days) after their first purchase. Formula: count(customers with ≥2 orders within window) / count(new customers in cohort).

• Inventory Turnover Days (DIO) Definition: Average days inventory is held before sale. Formula: DIO = Average Inventory / (COGS per day) = (Average Inventory) / (COGS / days_in_period). Alternatively, DIO = 365 / (COGS / Average Inventory). Notes: Average Inventory = (Beginning Inventory + Ending Inventory) / 2, unless a more granular average is available.

• Gross Margin (%) Definition: Percentage of revenue retained after cost of goods sold. Formula: (Net Revenue − COGS) / Net Revenue. Notes: Net Revenue is revenue net of discounts, returns, and taxes. COGS follows accounting policy (product cost plus inbound freight if applicable). Shipping/fulfillment costs are excluded unless defined in COGS by Finance.

Interactions and drill-downs

• Filters: Date range, market/region, channel/campaign, product category, customer segment, acquisition cohort, repurchase window, order vs revenue date.
• Drill paths:
  • GMV → category → product → order-level detail.
  • CAC → channel/campaign → creative/ad set → cost and first-order count.
  • LTV → cohort → segment (channel, product of first purchase) → monthly cashflows.
  • Repurchase → cohort → time-to-second-purchase distribution → drivers (category, price band).
  • DIO → category → SKU → aging buckets and stock cover.
  • Gross Margin → category/SKU → price, discount, cost variance.
• Tooltips: Calculation notes, data lineage tags, and data freshness indicators.

Alerts and targets (configurable)

• Alert if LTV:CAC ratio falls below threshold (e.g., 2.0) or CAC payback exceeds target months.
• Alert if DIO exceeds target by category (stock risk) or falls too low (stockout risk).
• Alert on margin compression beyond tolerance at category/SKU level.
• GMV underperformance vs plan by region/channel.

Data sources and refresh

• Commerce/Order system for orders, returns, discounts; Finance/GL for COGS and revenue recognition; Marketing/Ad platforms and CRM for spend and acquisition; ERP/WMS for inventory balances.
• Refresh cadence: Daily for commerce and marketing; daily or monthly close for COGS and inventory depending on accounting processes. Data freshness displayed in header.

Usage guidance (what to look for)

• Growth efficiency: Track LTV vs CAC trends and the LTV:CAC ratio. Investigate channels where CAC rises without corresponding LTV improvement.
• Retention health: Monitor Repurchase Rate by cohort; identify cohorts or products with declining second-purchase conversion and address onboarding or assortment gaps.
• Profitability: Watch Gross Margin alongside discounting and mix; drill into categories with margin erosion.
• Working capital: Use DIO by category/SKU to optimize buys and markdowns; rebalance inventory where DIO deviates from target.
• GMV quality: Compare GMV to Net Revenue and returns rate (available in tooltips) to ensure growth is translating to recognized revenue.

Governance and notes

• Metric ownership: Finance owns revenue, COGS, margin; Growth owns CAC; Analytics owns LTV and cohort logic; Supply Chain owns inventory metrics.
• All definitions and data lineage are documented in the dashboard’s info panel. Any deviations from defaults (e.g., inclusion of shipping in COGS) are flagged on-screen.