IMS databases use hierarchical data models where retrieval follows a specific sequence: top to bottom within the hierarchy, front to back among twins (segments with the same parent), and left to right across the tree structure. This deterministic order ensures predictable navigation through the database hierarchy.