Understanding the Mid‑Platform (中台) Concept: Background, Definition, Purpose, Features, and Goals

Background

The concept of “mid‑platform” (中台) derives from the US military’s command system, representing an efficient, flexible, and powerful command structure.

In December 2015 Alibaba upgraded its organization to a “big mid‑platform, small front‑end” model, aiming to break the traditional tree structure, bring the front‑end closer to users, and place support services in the mid‑platform as a platform‑support role.

Definition

“Mid‑platform” literally sits between the front‑end and back‑end.

The mid‑platform extracts common needs from various business lines, packages them as componentized resource bundles, and provides them to front‑end departments via interfaces, enabling more flexible product updates, faster innovation, and minimizing duplicate‑wheel KPI projects.

The front‑end (front‑office) requests resources directly from the shared services department; shared modules such as search, components, and data technologies are reused without modifying the underlying layer each time.

Purpose

Why does Alibaba establish a “big mid‑platform, small front‑end”?

Initially Alibaba had only a Taobao division, later adding Tmall. The same technical team supported both, leading to two siloed e‑commerce systems with overlapping functions (products, transactions, payments, logistics, etc.). To address this, Alibaba created a shared‑business division, consolidating common services into reusable components.

The mid‑platform is essentially a shared‑service architecture that abstracts common services (e.g., user, order) and specific business domains (e.g., education courses, instructors) into micro‑services, avoiding repeated investment.

Features

Service Reusability : Implements SOA principles; loosely coupled services enable business reuse, with the mid‑platform providing shared foundational capabilities to the front‑end.

Service Evolution : As new business lines join, shared services evolve from thin functions to robust, adaptable assets, becoming valuable IT resources.

Data Accumulation : Consolidated data across services fuels big‑data analysis.

Rapid Customer Response : Combines shared services quickly to meet new business needs.

Cost Reduction : New businesses reuse existing services, lowering personnel costs.

Development Efficiency : Developers focus on specific domains, accelerating development and maintenance.

Goals

First , decouple front‑end applications from back‑end services, extracting and consolidating capabilities such as order management, questionnaire management, product sales, and inventory.

Second , integrate duplicate or similar services, ensuring each service is generic enough to avoid “twin” services.

Finally , centralized service governance enhances integrated operations; each service must support multiple front‑ends, requiring stronger capabilities in security, backup, and monitoring.

Identifying a Mid‑Platform

A true mid‑platform solves the problem of “user‑centric continuous scalable innovation” by turning diverse back‑end resources into front‑end‑friendly capabilities.

1. Business Mid‑Platform : Provides reusable services such as user and order centers.

2. Data Mid‑Platform : Offers data analysis capabilities for rapid insight.

3. Mobile & Algorithm Mid‑Platform : Delivers personalized services, enhancing user experience.

4. Technical Mid‑Platform : Supplies infrastructure, distributed databases, and other foundational technologies.

5. R&D Mid‑Platform : Supports project setup, progress management, testing, CI/CD, acting as a rapid‑deployment base.

6. Organizational Mid‑Platform : Manages investment, risk, and resource scheduling, acting as the command center for the entire operation.

Ultimately, whether a platform qualifies as a mid‑platform is judged by the front‑end’s ability to leverage it for user‑centric innovation.

References

https://www.jianshu.com/p/a88b80d88284

https://www.iyiou.com/p/92012.html

https://yq.aliyun.com/articles/688225