In this follow-up article, we continue to explore AP functionalities focusing this time on the functionalities that simplify managing non-procurable spend.
If you missed the previous article on AP tips you can find it here. In addition, during our recent Basware Bespoke event Perttu Nihti, Chief Product Officer, discussed the product development strategy for our AP solution and in case you’re interested in hearing more, we have a recording available here.
1. Recurring Invoice Settings: Recognizing non-procurable spend
Most of the spend typically follows the purchase order process but there can also be a significant amount of spend that is non-procurable. This spend can be things you would not create a purchase order for such as rent, utilities, or subscriptions. We have introduced Spend Plans to automate the processing of this non-procurable spend. By creating Spend Plans, you can automate coding and approval of non-PO-based spend.
Spend Plans require implementation to set them up, but you can start by recognizing whether you have these kinds of invoices. We have a setting available that can help your users receive alerts when an invoice resembles an actual, recurring invoice that you could automate going forward.
You can set up the parameters to define how you want to recognize or set your pattern for these recurring invoices – options include the Gross Sum, plus or minus certain tolerances, invoice date plus tolerance, the supplier, etc.
Learn more about Recurring Invoice Settings
How to enable this feature? This feature can be enabled for each organization from the Data Management > Recurring invoice settings. Please contact Basware Support if you need help with the settings.
2. SmartCoding
The last functionality we’ll introduce is also related to the handling of non-PO spend. Basware SmartCoding is a machine-learning-based solution supporting your users with the coding of invoices that are not linked to a purchase order or a spend plan. By analyzing historical coding and invoice header data, it proposes coding suggestions based on individual dimensions, continually learning and improving with user validation. Users receive the top 5 coding suggestions, enhancing accuracy and efficiency. Users can decide when and where to generate coding, even routing low-value invoices with high-confidence coding directly through the workflow. 
The beauty of this service is that we can evaluate its success before you start. As this is a data-driven add-on service, we always run an initial analysis to see if SmartCoding suits your organization's needs. You then get a report that shows the anticipated value you can gain with SmartCoding.
For example, if in the initial analysis, we identify that you are always using an account code, cost center, and tax code, and we know that you have only one ERP running, we can then see, based on the analysis, that the expected accuracy is 90% for the different dimensions. So, before we enable the service for you, you will get a ready-made value proposition with the key metrics about the potential for taking SmartCoding into use, including the estimated time savings.
Another benefit of a machine-learning-based solution is that you don’t need to set up rules beforehand; this is exactly what we want to prevent. You will still get a coding proposal even if you have a new supplier. The solution looks at the fields in the invoice and predicts coding. The first time, the confidence level of the prediction may be around 40% or 50%, but you would have a proposal from day one. If you then change the proposed coding, you will have the opportunity to teach your system, and it will continue to learn by itself. Maintaining templates for this is unnecessary; it will learn as you do coding.
We are continuously developing this service and have recently introduced new functionalities into the service, like reading the invoice XML document and recognizing the potential need for split coding. Please read the feature descriptions for more information.
This advanced Basware SmartCoding is an add-on service, so if you are interested in learning more and getting an initial analysis done, please contact your Customer Success Manager or Account Executive.
