Sending an invoice to a customer requires some labor, cost, and time, but does not guarantee that the invoice will be paid. For example, someone must print an invoice, separate the copy going to the customer, stuff it in an envelope and mail it, which may then take several days to reach the customer, be routed through its mailroom, reach the accounts payable department, and be entered into the customer’s computer system (where the data may be scrambled due to keypunching errors). The invoice may even be lost at the customer site and never entered into the compute system for payment at all.
To avoid all of these issues, a company can use electronic data interchange (EDI). Under this approach, a company’s computer system automatically issues an electronic invoice that is set up in a standard format (as defined by an international standard setting organization) and transmits it to a third-party mainframe computer, where it is left in an electronic mailbox. The customer’s computer automatically polls this mailbox several times a day and extracts the electronic invoice information. Once received, the format is automatically translated into the invoice format used by the recipient’s computer and stored in the accounting system’s database for payment. At no time does anyone have to manually handle the data, which eliminates the risk of lost or erroneous invoicing data. This is an excellent approach for those companies that can afford to invest in setting up EDI with their customers, since it fully automates a number of invoicing steps, resulting in a high degree of efficiency and reliability.
There are several problems with EDI that keep most smaller companies from using it, especially if they have many low-volume customer accounts. The main problem is that it takes time and persuasion to get a customer to agree to use EDI as the basis for receiving invoices. This may take several trips to each customer, including time to send trial transmissions to the customer’s computer to ensure that the system works properly. To do this with a large number of low-volume customers is not cost-effective, so the practice is generally confined to companies with high-volume customers, involving a great many invoices, so the investment by both parties pays off fairly quickly.
The other problem is that the most efficient EDI systems require some automation. A standard EDI system requires one to manually enter all transactions, as well as manually extract them from the EDI mailbox and keypunch them into the receiving computer. To fully automate the system, a company must have its software engineers program an interface between the accounting computer system and the EDI system, which can be an expensive undertaking. Without the interface, an EDI system is really nothing more than an expensive fax machine. Thus, installing a fully operational EDI system is usually limited to transactions with high-volume customers and requires a considerable programming expense to achieve full automation.