This is the third blog post in a series on touchless processing. In the first post, I likened touchless processing to the parable of the six blind men and the elephant. Almost every AP Manager has heard the phrase, yet there is no consistent understanding of what constitutes real touch processing. The second and subsequent posts are an attempt to define the key functional components of touchless processing, and align them with emerging best practices. For instance, the second post discussed training the OCR. Rigorous training is an essential step towards achieving true touchless processing. It is time consuming, prone to exceptions, and never truly “done.” We recognized this problem early on, and that’s why we developed invoice “training-as-a-service.” When users of AP Express encounter a new invoice format, they simply click on the “submit for training” button and the invoice is routed to our team of trainers. We train and test the invoice, then push the invoice back to you for continued processing. No need to slow down and no loss of momentum.
The second step in touchless processing is to define a rule set. Touchless processing operates according to a set of best practice “rules” defined by the users of the system. Many of these rules already exist even when operating in a manual environment.
For instance, there are rules associated with approvals based on the value of the invoice. There are rules associated with how invoices are processed when someone is out of the office. There are rules that deal with exceptions in terms of pricing and quantity, or perhaps item match.
These rules (and more) need to be defined and instrumented within your AP automation solution. The process of defining and refining rules is necessarily iterative. If tolerances are set too low, every invoice becomes an exception. Too high, and you risk compromising the integrity of the approval process.
A robust AP automation solution allows rules to be easily tweaked and tuned to meet your specific requirements. The best solutions go beyond purely metric-based rules and incorporate fuzzy logic to address situations where Boolean logic is not possible. Like invoice training, rule sets are under constant supervision and prone to refinement based on changing business requirements. Still, a well developed rule set can dramatically reduce invoice processing times and deliver enormous efficiency gains to an AP team.
Next weeks post will review the role of analytics in touchless processing. As a professor once said, “If you can’t measure it, you can’t manage it.”