A Business Analyst’s End‑to‑End Journey Using a Data Middle Platform: From Issue Identification to Data‑Driven Solutions

Data middle platforms (data hubs) are defined in many ways; Alibaba’s methodology emphasizes “One Data, One Service”, aiming to support digital transformation and data value realization, with core capabilities split into data development, data governance, and data application.

关于数据中台的概念和定义很多。从建设方法论角度阿里提出了“One Data，One Service”。从建设意义的角度是为企业数字化转型、数据价值变现。从核心能力又分为数据研发、数据治理、数据应用等模块。

To illustrate these concepts, the article follows a business analyst (Xiaoming) at an e‑commerce platform who must explain a 30% Q1 sales drop for a product category. He first locates the relevant metric in the indicator system, identifies the “sales amount” metric within the transaction domain, and examines its definition and dimensions.

Existing Metric Analysis

Locate the declining sales metric and its domain.

Request access to the underlying source tables through the permission‑approval workflow, noting high security labeling and data lineage considerations.

Query the “channel sales” metric via the self‑service data platform, discovering that the Taobao channel is the main cause of the overall decline, but lacking finer‑grained indicators for deeper diagnosis.

Exploratory Analysis

Use the data map to find a user‑behavior table containing channel information, request its access, and perform ad‑hoc SQL analysis.

Find that product exposure and click rates on Taobao are stable, suggesting the issue lies in conversion or inventory, but no suitable table exists for inventory data, prompting a data‑development request.

Data Development

Data developers align on data definitions with Xiaoming.

They design, integrate, develop, test, monitor, and publish the required inventory table in Hive.

The data is then exported to downstream storage and exposed via a unified API service.

Operational maintenance ensures daily updates (T+1) of the data.

Visualization

After delivery, Xiaoming uses the indicator platform and self‑service query tool to confirm that zero‑stock items in the Taobao channel caused the sales drop.

He creates a visual report and shares it with management through the reporting platform.

Productization (Data Application)

To continuously monitor the issue, Xiaoming builds a supply‑chain decision‑support system that automatically detects inventory shortages, generates replenishment suggestions, and pushes them to the procurement system.

The article concludes by revisiting Alibaba’s “OneData, OneService” methodology, emphasizing data as a reusable asset, the importance of unified indicator systems, data maps, and API gateways for efficient data consumption and lifecycle management.

Appendix

The appendix lists the various modules of the data middle platform (e.g., data integration, data development, data governance, data quality, cost optimization, data security, data services, self‑service query, reporting, dashboards, visual analysis) and their primary user groups.