Organizations need quality information for control and coordination of their operations. Yet, despite the resources expended on data production, data often proves to be unreliable or inadequate when used in practice. To understand this puzzle, I conducted an ethnographic study of the production of a database system within one organization. Database systems provide the scaffolding to hold the data that are to be produced and consumed on a regular basis. I find that the creation of a database is a laborious exercise fraught with tensions and entailing both fears and desires. I examine these fears and desires from three separate lenses reflecting the various purposes that people imagine for the database – efficiency (rational lens), accountability (control lens), and comparability (standardization lens). As these purposes interact in the activities of database production, the fears and desires heighten to eventually produce an oversized database system with standardized, simplified, and abstracted data that undermines its stated purpose and generates significant difficulties for its users.