Monitor Data Quality in Context
Learning Objectives
After completing this unit, you’ll be able to:
- Explain why data quality depends on business purpose, persona needs, and permissions.
- Describe how user security context influences what should be monitored.
Review Data from Different Angles
Data quality can't be judged in isolation. A field that appears complete across an object might not be relevant to a particular business process, while a field that appears sparsely populated might be exactly what a specific team needs.
Many business applications, including Salesforce, allow a single object to contain hundreds of fields and records that support many different users and business processes. Which fields and records are relevant to a particular user is controlled through security settings, page layouts, sharing rules, and other configuration.
To understand whether data is truly fit for purpose, organizations often need to assess and monitor the same object from the perspective of different user personas. This helps determine which information is available, being adopted, and trusted by each group of users responsible for a business process.
Purpose-Driven Data Quality
Different teams use the same data in different ways.
- Service agents need current contact information to resolve cases quickly.
- Sales teams depend on qualification and account information to progress opportunities.
- Marketing teams need trusted contact points and segmentation attributes.
- AI agents require sufficient context to make accurate decisions while respecting the same security controls as the users they assist.
Monitoring must therefore align with the business purpose, not simply whether a field contains values.
Assess and Monitor Data in the User’s Security Context
Organizations can evaluate data quality from a specific user’s perspective by profiling data or running targeted queries using that user's security context. This reveals which records and fields are actually visible to the intended audience and whether those users have the information needed to perform their work.
However, visibility alone does not provide the complete picture. Understanding how data is used often requires combining several sources of information, including:
- User permissions and sharing rules
- Page layouts and user experience
- Reports and dashboards
- Automation and workflow dependencies
- Data profiling insights such as completeness, usage patterns, and value distributions
Together, these provide the context needed to determine whether data remains reliable and fit for purpose for a specific persona.
NTO Differentiates Consumer Versus Business Customer Needs
NTO supports both consumer and business customers using shared Salesforce objects. Customer service representatives, sales teams, and managers each interact with the same data, but through different permissions and user experiences.
As NTO prepares for its Case Deflection AI initiative, Luna wants to determine whether service agents have the information they need to resolve customer issues efficiently.
Her first option is to analyze replicated data in the enterprise data lake, where IT performs much of its technical reporting.
Instead, Luna chooses to profile the data natively using the security context of the service agent role. Because her goal is to monitor user adoption and data reliability, she wants the assessment to reflect exactly what service agents can see and update.
During her review, Luna notices that a recently introduced field has a 0% fill rate for service agents. After investigating, she discovers the field was never added to the service agent permission set. Although the field exists and contains data for other users, it was never available to the people expected to maintain it.
By monitoring data in the appropriate user context, Luna identifies and resolves the issue before it impacts reporting, customer service, or the upcoming AI implementation.
Unit Recap
Data quality depends on business context. Monitoring should evaluate data from the perspective of the people or AI agents who rely on it, taking into account their business purpose, permissions, and responsibilities. By combining security context with data profiling and metadata insights, organizations can identify meaningful risks and focus improvements where they have the greatest business impact.
In the next unit, you explore how to choose the right monitoring methods and cadence based on business risk, rate of change, and operational impact.