A well designed sales tax system must function off of one inventory item database, with a multi-jurisdictional state tax matrix in order to provide the proper audit trial to satisfy department of revenue compliance; which will minimize audit liability risk. Item taxability should lie in an algorithm that combines the item master, with the customer master for location, and the jurisdictional matrix for taxability. This type of design will allow a state auditor to obtain an item sales history by location. This process will minimize the risk of an auditor using alternative measures of averaging your data in order to compute liability during a sales tax audit; otherwise the auditor will use averaging methodologies if they determine they cannot rely on your system data. This may result in over payment of taxes in an audit due to the inability of a company to prove to the auditor otherwise; due to poor system design. In accounting we call this a robust point of sale system that provides sales item history by location. These types of systems apply best for retail, wholesale and manufacturing companies that sell tangible personal property.
However, the same logic will apply to a service organization that has revenue streams from software, service, implementation and maintenance type sales and contracts. In order to accommodate service type organizations, systems need to be augmented to accommodate different types of contract structures. Contract structures usually dictate how a jurisdiction will tax different type of sales of service organizations.