The Pros and Cons of the Middle‑Platform Model in Large Enterprises: A Data Department Perspective

First, the author clarifies that the piece is not aimed at Alibaba specifically but reflects a common phenomenon observed from a developer or user viewpoint, focusing on the data development side of the business.

The "middle platform" concept is not an internet or Alibaba invention; in large enterprises, any shared function relied upon by multiple departments can be extracted as an independent service platform, with examples such as Apple and Samsung.

Data departments are a classic example of such a platform, where the data platform’s surrounding capabilities and business characteristics give rise to various specialized middle platforms, like marketing‑data or supply‑chain‑data platforms.

Benefits of a unified service platform are numerous: centralised resources enable large‑scale projects, budgets scale with impact, and the platform can support both immediate business needs and long‑term research that may become valuable later.

For instance, data lineage—a capability rarely known ten years ago—can be centrally developed and rapidly iterated, replacing manual Excel tracking in many small companies.

Another advantage is avoiding duplicate development; a marketing‑data middle platform can provide common features such as coupons and red packets, allowing business teams to build on top of these foundations quickly and uniformly.

Successful experiences within one department can be replicated across others, facilitating knowledge transfer in large organisations where departmental walls are often deep.

However, the model also has downsides: the middle platform can be slow to respond, especially under Alibaba’s unique 361 assessment system, turning it into a bottleneck for business initiatives.

Excessive authority and bargaining power of the platform can lead to conflicts, with the P‑level hierarchy resembling a caste system and causing internal power struggles that hinder agile decision‑making.

Consequently, many companies are dismantling middle platforms to bring capabilities closer to the business, improving responsiveness.

Nevertheless, platform‑related capabilities remain essential; they are being re‑engineered to be more business‑centric, customizable, and agile, allowing each unit to tailor shared services to its specific needs.

In conclusion, the highly unified foundational abilities of data departments are well‑suited for customised development, and future enterprises will likely position these platform capabilities nearer to business lines, creating differentiated value that is valuable to showcase on resumes and in interviews.