One of my colleagues recently told me of an experience that has probably already happened to you : the receipt of an invoice (and a credit note, to boot!) for the amount of € 0.00. It happened recently to me with an Internet service provider of whom I was no longer a customer... But let's get back to this famous invoice :
« We thought we owed you € 0.00. But no, in fact, it is you who owe us € 0.00. »
« We cancel a credit note and charge you in your favor. »
These are essentially the messages conveyed by the invoice below, issued by an energy supplier. Would it be the gas bill that was used to cook the fish on April 1 ? Or would this provider have a small problem of data quality ?
« It's the computer's fault »
Who, following a contact with an operator, has not already received this terse reply in return to a question ? Sure enough, the processing is now automated. But one might be surprised that an accounting tool - or an ERP that generates invoices - produces unnecessary documents when the total reaches the sum of € 0.00. So, is this a question of common sense meeting the incompatibility of computer systems ? In any case, we urge this particular supplier to review their computerized data processing to avoid sending further unnecessary mail. A specialist in software testing and validation could have foreseen this not so exceptional hypothesis.
Improving the effectiveness of the invoice process, an impossible dream ?
In order to avoid (as much as possible) this scenario, it is necessary to identify inconsistencies at the data level and proceed with the cleaning of this data. « Data cleansing » is a significant option in the process of drastically reducing invoicing mistakes, whatever they may be, and an obligation that will soon be imposed by the GDPR (General Data Protection Regulation, see here for more information) if the data concerned are private and out-of-date.
In our example, the cost for the company is limited to the printing and sending of a mail - multiplied by the number of incorrect client data/accounts that cause automatic mail sending - but we can easily imagine from this scenario that part of the invoices never see their amount collected by the company for the same reason of lack of data quality (incomplete or incorrect mailing address, wrong customer name…). And the result is then far more serious for the company than the cost of unnecessary mail.
So, who's at fault here ? Employees responsible for the processing ? The computer program ? The database ?
Model Driven Data Engineering
At Rever, we have long been persuaded that the problem must be taken at its core : the information system must be of the highest quality to generate data of impeccable quality. The data quality is measured initially in relation to the data model that structured the application, it is then necessary to evaluate the capacity of the data model to meet the requirements of the users and/or the company.
Through MDDE (Model Driven Data Engineering, see more here), we operate at each of the three levels of quality measurement : data quality, database quality and data processing quality.