Digital Portraits of Data Governance: Measuring User Experience & Architecture

As the whole network enters the big data era, enterprises focus on using big data for refined marketing and operations. While many customer and employee profiling theories emerge, the underlying data governance lacks a comprehensive theoretical system. To avoid aimless governance work and quantify the contribution of data foundation, we need a "digital portrait" of data governance. This article discusses it from user experience and architecture quality perspectives.

01 User Experience Digital Portrait

Based on different perception angles, users are divided into external customers, internal users, management, and technical staff. For specific business scenarios, we depict the "technology empowerment" each group experiences.

1、External Customers

Functional Experience Indicators: These measure the usability and intuitiveness of the operating platform. By tracking clicks, page dwell time, and depth via various tracking points, we can discover frequently used functions, explore actual user needs, and focus on the smoothness and practicality of key features.

Platform Service Indicators: (1) API call rate reflects the activity level of outward data services. (2) The value added by data services to products can be measured, allocating a proportion of marketing and operation value improvements to data governance work, thus evaluating its empowerment effect.

2、Internal Users

Convenience: Previously, data requests were handled via email or administrative processes, lacking real‑time tracking and centralized management, causing high effort for staff. Automating and standardizing processes with online tools greatly improves convenience; the reduction ratio of manual tickets can serve as a metric.

Timeliness: Online data governance puts asset maps and standard architectures at users' fingertips. The end‑to‑end delivery time of key tasks reflects internal users' perception and can be measured by averaging node flow times.

Contribution: Beyond underlying data management, the output of data applications brings business value. Metrics such as BI tool usage and the number of models provided indicate user satisfaction with data‑driven outcomes.

3、Management

Quality Improvement: Management cares about the "clearness" of data warehouses and lakes. Regulatory reporting quality rates serve as a direct reflection of data governance effectiveness; DQC‑based indicators also represent data cleanliness for management.

Efficiency Improvement: Apart from regulatory requirements, data operation costs are crucial for overall management. Establishing standardized, efficient data architecture reduces reporting, storage, and operational costs, enabling refined operations and high‑efficiency profitability.

4、Technical Staff

Data Dictionary Scoring: When enterprises enforce strong development controls, the data dictionary acts like law in society; its logic must withstand scrutiny. Providing a scoring feedback mechanism on the dictionary query page guides designers, turns complaints into suggestions, and helps optimize the data dictionary experience.

02 Architecture Quality Digital Portrait

A unified data architecture should pursue high efficiency while reducing cost. Referring to the classic four paradigms of information architecture in "Huawei Data Way," we measure architecture enablement from model, distribution, standards, and assets.

1、Model

Public Layer Processing Frequency: The public layer stores fact and dimension data supporting top‑level metrics. Normalizing and consolidating dimensions improves reuse of public indicators and reduces duplicate processing; thus, the reuse rate of the public layer data model serves as an evaluation index.

Application Layer Reference Frequency: Similar to centrality algorithms in social networks, this index measures the systemic importance of data in the application layer, guiding asset inventory. Data lineage forms a directed, unweighted, acyclic graph. Frequently referenced assets usually originate from key business entities and are accessed across departments, helping identify "orphan" or temporary tables and reduce resource waste.

2、Distribution

Data Coverage: Large banks have hundreds of systems and thousands of tables distributed nationwide. Data collection is the first step of asset inventory; measuring collection coverage across all systems clarifies progress and identifies uncovered data sources.

Data Redundancy: Redundancy refers to duplicate data within the same layer, arising from multiple physical locations storing identical meaning data or from architectural models with overlapping components.

Data Volume: Data volume describes the overall size of the data middle‑platform, including absolute amount and growth proportion over time. The optimal volume depends on the bank's specific context and must be evaluated dialectically.

3、Standards

Standard Stability: Data standards normalize meanings and structures, requiring consistency and non‑overlapping definitions to avoid "data conflicts".

Standard Landing Rate: Assuming complete technical specifications and authoritative publication, the landing rate reflects execution of standards in the "last mile." Automated tools can calculate landing rates for various layers, intelligently uncovering potential issues.

4、Assets

Technical Metadata Statistics: Technical metadata links source data to the data warehouse, recording the lifecycle from creation to retirement. Selected metrics such as system coverage, table‑level coverage, field name validity, and enumeration validity represent the output benefits of technical assets.

Enterprise Activity Hit Rate: Data assets are digital representations extracted from business processes and models. Higher hit rates of label assets on business actions, indicator assets on report statistics, and user access volume of report assets indicate greater accuracy of asset content mapping to enterprise activities.

04 Final Thoughts

With the deepening of enterprise digital transformation, the "digital portrait of data governance" will become more refined in methodology and practice, enhancing content value, security performance, and user experience. Dynamically measuring data governance effectiveness and establishing a tailored "North Star" metric is essential for any company in the smart transformation stage, and its success will generate immeasurable commercial value.