Accounts Payable Vendor Invoice Processing An Automation Journey
A well-executed business strategy brings along a rapid growth and accelerated production; Furthermore, an increase in staff and customers.
A well-executed business strategy brings along a rapid growth and accelerated production; Furthermore, an increase in staff and customers. This also leads to massive volumes of invoice processing. According to a survey by IOFM (Insti-tute of Finance & Management) in 2019, the average number of invoices processed shot up 28% on a yearly basis.
One of Calpion’s customers, a large company in the catering and hospitality sector providing products and services for the travel industry, with an increasing number of customers, was scaling up exponentially to meet the demand. But their Accounts Payable was a complex process with manual data sources, tactical solutions, and limited resources impeding the sustainability of operations. According to a research by APQC (American Productivity & Quality Centre), 62% of total Accounts Payable costs are due to labor. Data errors in Purchase Orders and Invoices require manual intervention. Lack of a standardized Invoice Management process and increased turnaround time led to unpaid Invoices and delayed payments indicating significant nefficiencies.
The customer was processing invoices the traditional way. Paper-based invoices from suppliers and vendors were processed and scheduled for payments.
There were several moving parts in the process cycle. The customer received approximately 17,000 invoices in a month and over 550 invoices daily. Basis the Service Level Agreement (SLA) defined; two full-time employees (FTEs) processed the invoices four times a day. The invoices were categorized into ‘Good’, ‘Fair’, and ‘Bad’ buckets depending on their quality. Furthermore, this formed the base to define the level of exception as ‘High’, ‘Medium’,and ‘Low’.
Each day, the two FTEs would retrieve the Invoice Information from a PDF file that was received via emails in Outlook or Web Form and convert the information to an XML document using Microsoft Excel. Data was rather unstructured and entered into an output file manually, which was then fed into a database. The completion time for each invoice was eight minutes.
The as-is process steps have been encapsulated below,
The customer received both paper and digitized invoices. Paper Invoices sent via Postal Service and other handwritten, illegible invoices were uploaded to Web Portal. This, in turn, uploaded the PDF to a shared location with MSC code, and invoice number added. The employees accessed the invoices placed in the shared location, read and processed 11 key fields as outlined in the SLA, including the identification of Cost center and General Ledger account such that costs were apportioned to the correct department. This processed invoice was then input as a unique row into the output XML file.
Similarly, Invoices received as PDF via Outlook were uploaded into the shared location where Database marked an entry of each record and outputs into the said XML file.
The FTE also spent considerable time sorting unprocessed invoices often resulting in duplicated processing or sorting duplicated invoices sent by the vendor.
The scalability of this process manually would be time-consuming, likely to introduce more inefficiency, and hinder business sustainability. Besides,manual scaling of the process would mean hiring additional workforceeventually, leading to cost overruns. The objective was to do more,efficiently, and without escalating operational costs to the firm. Any furtherdelays in the process would directly impact business operations.
The client was keen on running as lean as possible and realizing thatinvestment in RPA was crucial for their digital transformation journey. Since RPA bots have the ability to automate the input of data, convertun structured data elements to structured data, reconcile errors, and reduce manual intervention for exception handling.
Thus, Calpion chose Ui Path, known for its easy to use interface and robust customer support. The Calpion team was instrumental in planning and designing the RPA Transformation Roadmap for the customer. Unattended bots were built as an end to end solution.
All digitized invoices sent by vendor and scanned Paper invoices was pushed through the email client. A user interface was also built to upload handwritten and paper invoices. The bot continuously refreshed and read mailbox for incoming invoices. Once an invoice was found, the bot downloaded the attachment into a shared folder, which in turn would create a unique record in the database. Duplicated invoices were filtered out at this stage.
The data was extracted from the invoice through a straight through processing (STP) however, if an exception was encountered, the same would be flagged to the end user. The End User reviewed each exception,extracting relevant fields as outlined in the SLA.
The user also uploaded the Invoice Number, MSC Code, alongside the scanned invoice. The bot would read these for further processing based on a mapping file named ‘Master Invoice BOT’. This mapping table comprised of the related words to fetch key fields outlined in SLA. Any unhandled PDF would show up as an exception for the FTE to process manually. A user inter-face with login credentials and privileges was enabled.
If the bot was unable to process a data element or if crucial data points are missing in the invoice, an error code was generated, and an exception was created for the FTE to process the field manually. A Data entry page to input such fields was created. Only authorized users could further process such invoices. Upon data entry, the bot would validate input data and generate an XML file and save it in the shared folder.
All exceptions are consolidated and generated via a Report after the execution of each batch, four times a day.
Invoice is now uploaded into the Accounts Payable ERP where the uploaded invoice was sent for further approvals and finally the payment being released.
Furthermore, the user interface enabled a user to create and export adhoc reports using customized filters.
A log was also created during the data extraction process for audit purposes. The log would reflect the number of bots that were involved in the process, number of transactions and invoices processed, error logs containing number of errors, and information on each of them.
The solution was monitored post-go-live, and the following business outcomes were accomplished:
The client was able to scale up its invoice processing from a 8/5 to a 24/7 process.
Automated Invoice Processing streamlined the operations, and hence turn around time was reduced. Processing time was reduced from 8 to 0.83minutes resulting in 90% savings.
No backlogs since all invoices were processed on the same day.Processing time in AP was reduced by 10 days which allowed for extended analysis of costing trends, reduction in accruals, prompt pay discounting review, analysis for inventory management, and meeting any prerequisite payment terms.
10% human effort required only to handle exception cases and high-value work such as MI analysis and deep-dive leading to a boost in employee morale and a productivity boost by 65%
0% error rate in data from PDF and SAP thus minimizing operational risk and data breaches.
Improved Communication, transparency and ease of data accessibility rendering better customer service and vendor relationship.
Calpion, Inc. is a software development and innovation powerhouse headquartered in Dallas, TX. With over 16 yrs. of providing cutting-edge software development and testing services from its 3 Global Delivery locations, Calpion today is at the forefront of innovating and delivering Robotic Process Automation (RPA) and Machine Learning (ML) to its global list of Customers. With its partnership with industry leaders in the RPA space, deep domain knowledge and process expertise and in-house software development & implementation specialists, Calpion endeavors to deliver cutting-edge RPA processes that can automate finance tasks, reduce labor and increase efficiency.