Skip To Content

Data quality and validation

Available with Data Reviewer license.

Data used in your GIS projects for visualization, analysis, compilation, and sharing should meet a defined standard for quality. Data quality requirements vary from project to project. Requirements for how accurate or complete a dataset needs to be are based on how the data will be used. Requirements are also influenced by technical, product, and client requirements. ArcGIS Pro has a data quality management system that provides tools to systematically perform quality assurance and quality control (QA/QC) on your GIS data, efficiently work through the issues identified, and generate quality-related reports.

Validate your data

Data validation is an iterative process that uses formal methods of evaluating a dataset's adherence to a defined quality standard. ArcGIS Pro provides you with tools for automating and simplifying data quality assurance (QA) and quality control (QC) through automated or semi-automated workflows. These tools can help detect anomalies with features, attributes, and spatial relationships in your data.

Automated review

Automated review is the ability to evaluate a feature's quality without human intervention. This includes capabilities such as Reviewer checks that ensure data integrity by performing spatial, geometric, and attribute validation. Checks are configured to validate data based on specific conditions. Some checks search for conditions, such as polygon slivers or cutbacks, while others search for features with specific spatial relationships. For more information, see ArcGIS Data Reviewer checks in ArcGIS Pro.

Semi-automated review

Semi-automated review methods are guided workflows that require some form of human interaction and input. Examples include visual review workflows that are employed to discover missing, misplaced or miscoded features. Data Reviewer provides a series of simple-to-use tools to streamline workflows required to detect errors that cannot be found using automated methods. For more information, see Identify errors on existing features and Identify missing features.

Compile the results

Results from data validation are organized into sessions. Sessions represent validation and quality control transactions performed by data checks or manual review. Sessions can be identified by user-defined names and are stored in the Reviewer workspace. The Reviewer workspace can be stored in a file or enterprise geodatabase and must contain a Reviewer schema.


Results represent an object in your GIS, such as a feature, table row, or metadata element, that has been marked as an anomaly by validation (using data checks) or manual inspection. Results include information about the source of the error, detection method, date and time, and life cycle phase. Results can have geometry. This identifies a feature (or a portion of a feature) that has been classified as an anomaly (or error) by a data check. This geometry enables you to quickly navigate to the specific area in which the error was detected. Checks run against tables will create results without geometries.

You can apply a severity rating to results to indicate importance. Severity is set within a data check and applies to results. Severity is represented on a scale of values, from 1 (highest) to 5 (lowest). For instance, to indicate that buildings found on lakes are a high priority to review and fix, you can assign the check a severity rating of 1. Any result created by this check will have a severity of 1.

Track life cycle status

You can keep track of a result's life cycle within the QA/QC process. There are three life cycle phases: Review, Correction, and Verification.

Reviewer life cycle statuses

Each phase has status values that describe how the result was handled. A result progresses through each phase as it is reviewed, corrected, and finally verified. Life cycle phases describe the who, when, and what in the process of correcting and verifying the error results.


The Review phase is the starting point for all results. A result's reviewed status value is based on the tool that submitted it.

A result with a status of Reviewed means that it needs correction. A result with a status of Unacceptable indicates that it has been returned for correction after review during the Verification phase.


The next step in a result's life cycle is the Correction phase. In this phase, the result submitted by one of the methods above is addressed. This usually involves a change to feature geometry or attribution. Once the feature has been fixed, a correction status can be applied to the result.

A result with a correction status of Resolved means that it is ready for verification. A feature result with a status of Marked As Exception indicates that it has been reviewed and is recommended to be an exception to the automated check that generated the result.


The last stage in a result's life cycle is the Verification phase. In this step, the fix to the result is verified. A verification status is applied to the result. A result with a Verified status of Acceptable means that it has completed the QA/QC workflow. A feature result with a verified status of Exception indicates that it is an exception to the automated check that generated the result.

The result identified by the data check or manual review has been reviewed, corrected, and verified.

Determine data quality

Knowing the level of quality of your GIS data is important for analysis and decision making. Generating reports can provide you with information regarding the types of validation executed, what was validated, and the number of results identified. If these figures are in the acceptable range, you can feel confident using the data for your GIS projects.