Enterprise Data Asset Quality: A Management-Standard Conformity-Benefit Realization Framework and Formation Mechanisms

Beyond Technical Accuracy: Mapping the Lifecycle of Data Value

Why do some companies treat data as a strategic engine while others treat it as a mounting technical debt? As organizations transition from simple data collection to treating information as a core economic asset, they face a widening gap between having "big data" and having "high-quality data." Researchers have recently developed a new way to measure how good a company's data assets truly are. They found that high-quality data isn't just about technical accuracy. It also requires strong management and the ability to turn that data into real business value.

The difficulty lies in the fact that traditional data quality metrics only tell half the story. Engineers typically check if a database is complete or accurate. While these are vital for technical reliability, the authors argue this perspective fails to capture the reality of data as an asset. An asset must be owned, controlled, and capable of generating sustained economic benefits.

The disconnect between quality and utility

Current approaches to data assessment often suffer from a narrow focus on intrinsic technical attributes. This means focusing only on the data itself rather than how it functions within a business. A perfectly accurate dataset is useless if no one in the organization has the authority to use it. It is also useless if it does not align with any strategic goals.

The paper notes that existing frameworks often specialize in single dimensions. Some focus on normative requirements (compliance), while others focus on diagnostic maturity (capability). This creates a fragmented landscape. A company might have technically perfect data that nonetheless fails to support any strategic decision-making. Because existing studies often lack a unified definition, enterprises struggle to understand why their massive data investments are not yielding a return.

A three-dimensional engine for value

To solve this, the authors propose a multidimensional framework. They treat data asset quality as a coupled system. They model it as a chain reaction consisting of three core pillars: Data Asset Management Capability (Management), Data Quality Standard Conformity (Standard), and Data Asset Benefit Realization Capability (Benefit).

The mechanism operates as a cascading sequence: 1. Management as the Foundation: This involves institutional arrangements. These include organizational structures and resource allocation that ensure data is handled consistently. 2. Standard as the Enhancer: Effective management allows a firm to implement rigorous standards. This ensures data is accurate, consistent, and accessible. It turns raw information into something interpretable and transferable. 3. Benefit as the Realization: Once data is standardized, it can be embedded into business processes. This produces observable outcomes, such as increased revenue, cost savings, or improved decision support.

The authors visualize this structural logic in .

This diagram illustrates how these three dimensions interact to form a cohesive quality profile.

Evidence from the field

The researchers validated this framework using a massive dataset. This included 1,500 questionnaires from enterprises in Hubei Province. The sectors studied included healthcare, technology, and urban governance. The paper reports an overall mean data asset quality score of 3.82. This score suggests that the surveyed enterprises are at an intermediate level of development.

One of the most striking findings concerns the "chained" nature of the mechanism. Using Partial Least Squares Structural Equation Modeling (PLS-SEM)—a method used to estimate linear relationships between variables—the authors report a powerful connection. Management Capability has a direct effect on Standard Conformity (coefficient of 0.885). It also has a significant direct effect on Benefit Realization (0.633). This suggests that governance is the primary engine driving both technical standards and eventual value.

The study also reveals significant industry-specific variance. As shown in, the healthcare sector achieves the highest quality score (4.62).

This is likely due to highly regulated environments that mandate strict standardization. Conversely, the urban governance sector reports the lowest score (2.86). Interestingly, shows that in almost all sectors, management scores lag behind standard and benefit scores.

This indicates a widespread "application bias." Firms prioritize immediate data usage over the long-term governance required to sustain it.

Identifying the bottlenecks

The authors do not just say management is important. They use Necessary Condition Analysis (NCA)—a method used to identify prerequisites for an outcome—to prove it is a requirement. Think of a necessary condition like oxygen for fire. You can have heat, fuel, and a spark, but without oxygen, you get no flame. The paper finds that Management, Standard, and Benefit are all "very strong" necessary conditions. If any one of these is low, high-quality data assets are mathematically impossible to achieve.

However, the study also acknowledges several complexities. The findings are primarily drawn from enterprises in Hubei Province. This may limit how easily these results apply to different regions. Furthermore, because the data relies on self-reported surveys, there is an inherent risk of response bias. Companies may overestimate their own governance maturity. Finally, the model is a snapshot in time. It does not account for how data quality might evolve as a company grows.

The verdict: Build the foundation first

If you are looking to deploy a data strategy, the verdict is clear: don't skip the boring parts.

The research demonstrates that you cannot "shortcut" to value. Many organizations attempt to jump straight to the "Benefit" stage by deploying flashy analytics tools. However, the authors show that without a "Management" foundation, these efforts will hit a ceiling. The strongest path to success is the "management foundation $\rightarrow$ standard enhancement $\rightarrow$ value realization" chain.

For practitioners, the takeaway is to prioritize governance, strategy, and architecture. Do not focus solely on the latest AI application. If your management capability is weak, your standards will be inconsistent. Eventually, your "valuable" data will become a liability.

Figures from the paper