What Skills Do Data Product Managers Need in a Data Middle Platform?

Definition of a Data Middle Platform

A data middle platform aggregates and harmonises all internal (and optionally external) data of an enterprise, eliminating data silos and reconciling inconsistent standards. By providing a single source of truth, it enables unified data‑service interfaces for downstream applications.

Typical illustration: three systems define a field called “bun” differently – A: bun with skin and filling; B: bun with meat and vegetables; C: bun that satisfies hunger. The platform must standardise this definition so that downstream queries refer to the same entity.

Another example shows a user “Stone” with activity logs across a rental app, a job‑search app, and a food‑delivery app. By linking these logs under a unified identity (One ID), the platform can infer that Stone is looking for a house in Haidian while currently residing in Chaoyang, enabling richer user‑profile construction.

Value and Emerging Challenges

After establishing data governance, the platform serves as a unified data‑service gateway, continuously enriching the data ecosystem for scenarios such as One Data, One ID, and One Service. Technically, the stack resembles a combination of a data warehouse and middleware, requiring distributed storage and compute (e.g., Hadoop, Spark) and neutral data‑service APIs.

The massive aggregation of personal data raises privacy concerns: who controls the “lock” on user information, and how are access permissions enforced?

Key Product‑Manager Roles

Data Product Manager (Rule‑Maker)

Primary duties include designing a coherent data model, standardising definitions, and driving cross‑department adoption of those standards. The role abstracts business problems into reusable data products rather than answering ad‑hoc analytical queries.

Technical competencies :

Programming / query languages: SQL, Python, R Big‑data ecosystem: familiarity with Hadoop clusters, Spark, data‑mining techniques

Data‑visualisation tools: Tableau, FineBI, Cogons

Soft competencies :

Effective communication across business units lacking direct KPI incentives

Strategic thinking and industry insight

Logical reasoning for data‑model design and extension

Data Platform Product Manager (Ark Builder)

This role focuses on building the platform components that enable data products. Responsibilities include designing and delivering scheduling systems, real‑time processing pipelines, Hadoop‑ecosystem services, BI dashboards (including mobile BI), and user‑profile platforms.

Key abilities:

Understanding of data‑visualisation and overall platform architecture

Proficiency in SQL and awareness of the data lifecycle (ingestion → storage → processing → serving)

Design of data‑access permission models and governance controls

Vision for commercialising the platform and integrating it with downstream products

Skill Summary for Both Roles

Data Product Manager : data modelling, metric definition, data‑pipeline awareness, visualization, Hadoop/Spark knowledge, strong stakeholder negotiation.

Data Platform Product Manager : platform tooling (scheduler, streaming, batch), API design, permission & governance, SQL, visualization concepts, product‑market fit planning.

Both positions require a blend of technical depth and product sense, with the Data Product Manager leaning toward data‑model governance and the Data Platform Product Manager concentrating on the infrastructure that delivers those models to users.