While automated review is an efficient tool for checking data, it is limited in some ways. For the most part, automated review validates data that already exists and it is only able to validate data against those rules that can be defined and have been implemented. Generally, automated review is not checking for data that does not exist (features that are missing from a dataset) and it is not always possible to identify features with incorrect locations or attributes unless those characteristics are in violation of one of the business rules.
Data Reviewer provides a number of ways to simplify, organize, and structure the visual review process and provides tools to write those errors to the Reviewer workspace. These include web-based tools for gathering feedback from non-GIS users of map services, identifying missing or incorrect features with ArcGIS for Desktop tools, and random sampling of features to facilitate visual review of a statistically valid subset of features.