Digital Portrait of Data Governance: Measuring User Experience and Architecture Quality

Introduction

As enterprises enter the big‑data era, attention shifts to using data for refined marketing and operations. While many customer and employee portraits emerge, the underlying data governance lacks a systematic theory. To avoid aimless governance work and to quantify its contribution, a digital portrait of data governance is proposed, examined from user‑experience and architecture‑quality perspectives.

01 User‑Experience Digital Portrait

Users are divided into four categories—external customers, internal users, management, and technical staff—each with specific business scenarios.

1. External Customers

Functional experience metrics : Track click behavior, page dwell time, and depth via event‑tracking to identify frequently used functions, enabling function redesign, peer‑product comparison, and user‑feedback surveys. Platform service metrics : Measure API call rate to gauge external service activity, and assess product value uplift contributed by data services, allocating a proportion of marketing/operation gains back to data‑governance efforts.

2. Internal Users

Convenience : Replace manual email or admin‑process data requests with automated, standardized online tools; the reduction ratio of manual tickets serves as a convenience indicator. Timeliness : Record end‑to‑end delivery time of key tasks across the data‑governance workflow; average node turnover time reflects timeliness. Contribution : Metrics such as BI tool usage and model provision count indicate the business value delivered by data applications.

3. Management

Quality improvement : Data‑warehouse and data‑lake cleanliness is critical; data‑quality compliance rate (e.g., regulatory reporting standards) directly reflects governance effectiveness, supplemented by DQC‑based indicators. Efficiency improvement : Standardized, efficient data architecture reduces reporting workload, storage, and operational costs, supporting fine‑grained operations and profitability.

4. Technical Staff

Data‑dictionary rating : Implement a rating feedback mechanism on the data‑dictionary query page; guided prompts turn complaints into suggestions, improving the developer experience.

02 Architecture‑Quality Digital Portrait

Following the four classic paradigms of information architecture from “Huawei Data Way,” the architecture is evaluated on model, distribution, standards, and assets.

1. Model

Public‑layer processing frequency : Normalizing and consolidating dimensions in the public layer improves metric reuse and reduces duplicate processing; reuse rate serves as an evaluation index.

Application‑layer reference frequency : Analogous to a core‑person algorithm in social networks, this metric measures how often data assets are referenced across applications, guiding asset inventory and identifying “orphan” or temporary tables to cut storage costs.

2. Distribution

Data coverage : For large banks with hundreds of systems and thousands of tables, measuring the coverage ratio of collected data against the total system landscape indicates collection progress and uncovers missing sources.

Data redundancy : Redundancy exists when identical data resides in multiple physical locations or when the architectural model contains overlapping components.

Data volume : Captures the absolute size of the data‑mid‑platform and its growth rate; the ideal level depends on the bank’s context.

3. Standards

Standard stability : Standards must ensure unified definitions without cross‑conflicts, preventing “data fights.”

Standard‑coverage rate : Assuming complete technical specifications and authoritative publication, this rate reflects execution of standards in the “last mile,” calculated via automated tools that slice data by layers.

4. Assets

Technical metadata statistics : Technical metadata links source data to the data warehouse, recording the lifecycle of data. Indicators such as system coverage, table‑level coverage, field‑name validity, and enumeration validity quantify technical asset output.

Enterprise‑activity hit rate : Tags, indicator assets, and report assets are evaluated by their hit rates on business processes, reports, and user accesses; higher rates indicate better alignment of data assets with enterprise activities.

Conclusion

As digital transformation deepens, the digital portrait of data governance will mature from methodology to practice, enhancing content value, security, and user experience. Dynamically measuring governance effectiveness and establishing a “North Star” metric are essential for any organization in the smart‑transformation stage, promising immeasurable commercial value.