Siebel installation populates seed data (reference data like picklists, LOVs) and repository data (metadata, schema definitions, object definitions). User data is created post-installation through application use, not during the installation process.