数据源列表生成器

Below are five credible data sources suitable for monthly monitoring of the new energy passenger vehicle (NEPV; BEV/PHEV/EREV) market. Each includes scope, cadence, and practical notes for integration.

1. China Passenger Car Association (CPCA)

• Scope: China retail and wholesale NEV passenger car volumes; OEM- and (in subscription) model-level breakdowns; market share, inventory indicators.
• Cadence: Monthly.
• Use: Core source for China market tracking (retail vs. wholesale), model mix, and competitive dynamics.
• Notes: “NEV” definition aligns with China policy (BEV, PHEV, EREV, fuel-cell). Public summaries are free; detailed datasets require subscription.
• Link: http://www.cpcaauto.com/

2. European Alternative Fuels Observatory (EAFO)

• Scope: EU-focused BEV/PHEV monthly new registrations by country; in-use stock; charging point counts (public/fast), connectors, and rollout trends.
• Cadence: Monthly/near-monthly updates.
• Use: EU market volumes and infrastructure capacity monitoring; supports market-share tracking and charging supply-demand analyses.
• Notes: Harmonized EU methodology; verify latest refresh dates per country. Good complement to national agencies (e.g., KBA, DGT).
• Link: https://alternative-fuels-observatory.ec.europa.eu/

3. Argonne National Laboratory (US DOE) – Light-Duty Plug-in EV Sales (U.S.)

• Scope: Monthly U.S. BEV/PHEV sales estimates, market share, cumulative stock, and powertrain split for light-duty vehicles.
• Cadence: Monthly.
• Use: Authoritative U.S. EV volume tracking; consistent methodology over time; suitable for time-series dashboards and YoY/ MoM analysis.
• Notes: Based on industry sources and Argonne methods; treats sales (not registrations). Model-level detail is limited in public outputs.
• Link: https://www.anl.gov/es/light-duty-electric-vehicle-sales-in-the-united-states

4. EV-Volumes (Commercial)

• Scope: Global plug-in (BEV+PHEV) sales by country and model, market shares, and historical time series; selected pricing/incentive context.
• Cadence: Monthly updates (subscription).
• Use: Cross-country comparability; model-level benchmarking and share-of-segment analyses; global rollups.
• Notes: Paid access; ensure alignment of fuel-type taxonomy and reporting cut-off dates when merging with open sources.
• Link: https://www.ev-volumes.com/

5. S&P Global Mobility (IHS Markit) – New Registrations/Parc (Commercial)

• Scope: Global monthly new registrations (often VIN-based) with fuel type, model/trim, regional granularity; parc in operation; forecasting modules.
• Cadence: Monthly (with typical lags by country).
• Use: High-granularity competitive intelligence, mix-shift analyses, and funnel conversion (registration-based) vs. sales-based series reconciliation.
• Notes: Paid access; registration vs. sales differences require harmonization; country-specific lag structures should be modeled explicitly.
• Link: https://www.spglobal.com/mobility/en.html

Implementation notes for monthly monitoring:

• Normalize definitions: Align NEV/PEV taxonomies across sources (BEV, PHEV, EREV), and clearly separate sales vs. registrations.
• Handle lags: Model reporting lags by market (e.g., registration data may lag sales by weeks); apply nowcasting where needed.
• Granularity: Use CPCA for China retail detail, EAFO for EU-wide comparability, Argonne for U.S. national trends, and EV-Volumes/S&P for global/model-level fill.
• Quality control: Reconcile overlapping sources using priority rules and maintain a data dictionary with source-specific caveats.

以下为研究移动应用用户留存与渠道归因的5类核心数据源（含典型提供方、可用字段与常见用途）：

1. 应用内行为分析事件数据（如 Firebase Analytics/Google Analytics for Firebase、Amplitude、Mixpanel）

• 关键字段：user_id 或 app_instance_id、event_name、event_timestamp、session_id、device/OS/app_version、geo、用户属性（首日/首源标记）、UTM/归因标签（若已回填）
• 用途：构建留存与回访队列（D1/D7/D30 等）、会话与功能使用行为分析、分群与漏斗、版本或地域维度的留存差异评估

2. 移动归因平台（MMP）数据（如 AppsFlyer、Adjust、Branch、Singular）

• 关键字段：install_id/attribution_id、媒体渠道/广告系列/广告组/素材、触点类型与时间（click/view）、ATT 同意状态（iOS）、IDFA/GAID（在合规前提下）、再归因/再参与标记、SKAdNetwork 字段（conversion_value、postback 窗口）、成本回传
• 用途：安装与再参与的渠道归因、跨媒体去重、投放 ROI 与 LTV 按渠道/活动拆解、与应用内事件拼接以进行渠道维度留存与漏斗分析

3. 广告平台与隐私归因框架数据（如 Meta Ads、Google Ads、TikTok Ads；iOS SKAdNetwork；Android Privacy Sandbox Attribution Reporting）

• 关键字段：展示/点击/花费/转化（平台侧定义）、活动与素材层级维度；SKAdNetwork postback（coarse/fine conversion value、postback 窗口、源标识）；Android Attribution Reporting 事件级与聚合报告
• 用途：获取真实成本与曝光触达、在隐私框架下获取聚合转化信号、与 MMP/应用内事件对账与校准、支持 MMM（营销混合模型）和增量性评估

4. 应用商店与平台侧数据（Apple App Store Connect、Google Play Console 与 Install Referrer API）

• 关键字段：安装/App Units、商店内来源类型（如搜索/浏览/引用）、产品页浏览、转化率、订阅与续费指标；Google Play Install Referrer（referrer 字符串、点击/安装时间戳、UTM 参数）
• 用途：获客来源与商店转化诊断、与归因安装量对齐、订阅留存与流失监测、基于 Referrer 的安卓安装归因与反作弊信号增强

5. 服务器端业务日志与客户运营/用户数据平台（后端事件流与数据仓库；Braze、Airship、MoEngage 等）

• 关键字段：user_id、账户创建/活跃/注销时间、交易与收入事件、心跳或会话心跳、消息触达/打开（Push/Email/短信）、分群标签
• 用途：留存与回访的“真实活跃”口径、收入与 LTV 计算、生命周期与运营触达对留存的影响评估、与归因数据拼接进行渠道层级的营收与留存闭环

说明与实践要点：

• 在 iOS 上受 ATT 与 SKAdNetwork 约束，单用户级归因受限，应采用聚合分析与实验设计（增量测试、地理分区）作为补充。
• 数据拼接常用键包括 install_id/app_instance_id、install_timestamp、campaign_id，以及在合规前提下的广告标识符；需做好时间窗与去重策略。
• 统一事件与维度字典（渠道命名、活动层级、事件标准化）可显著降低归因与留存分析的偏差与对账成本。