Building a Data Middle Platform Indicator System for the Automotive Industry

In the automotive industry, data has become essential for strategy, product planning, and brand decisions, making a robust indicator system crucial for real‑time monitoring, decision support, and problem identification.

The speaker, Zhao Song, a senior big‑data product leader from a high‑end electric vehicle brand, shares a methodology for constructing such an indicator system.

Industry Pain Points : siloed "chimney" systems, inconsistent technology stacks, and fragmented data hinder business development and analysis efficiency, leading to six major symptoms (fragmentation, disorder, heaviness, slowness, lack, difficulty).

Indicator System as the Value Engine : The data middle platform evolves through four stages—database (OLTP), data warehouse (OLAP), data platform (Hadoop), and data middle platform (Alibaba’s concept)—enabling data assets to become valuable assets.

The platform comprises four layers: business applications (marketing, lead conversion, delivery, after‑sales), data products (decision support, CDP, data‑driven services), data governance (OneID, OneData, OneService), and data sources (customer, marketing, after‑sales data).

Indicator System Construction Method :

1. Clarify business goals across the full customer lifecycle (awareness, lead, conversion, referral, service).

2. Design indicator schemes using the AARRR (pirate) model, defining metrics, dimensions, and data sources.

3. Develop indicators (atomic, derived, composite) and implement real‑time (DataHub+Flink+Hologres) and offline (MaxCompute+Hologres) pipelines.

4. Manage indicators through governance processes: collaborative creation, naming standards, online dictionaries, metadata binding, and output for data products.

Practical applications include mobile dashboards, PC dashboards, and decision‑making screens.

Case Studies :

Case 1: Full‑cycle sales analysis for dealers and regions, using metrics such as test‑drive rate, card‑creation rate, and customer retention to pinpoint weak links and optimize strategies.

Case 2: Digital sand‑table for market insight, building competitor and self‑product indicators to improve market share and sales performance.

The presentation concludes with thanks to the audience.