From spotting trends to sporting shelves, do more with AI

The retail sector is in a state of transformation. The entire sector is attempting to cope with fast-changing customer shopping habits and the shift of emphasis from the high-street to the Web. Most retailers have already invested in and innovated with a variety of technologies including Artificial Intelligence, robotics, logistics automation, data analytics and self-service technologies in an effort to become more competitive, more customer-centric and more responsive to demand and opportunity. From our exclusive report, learn more about the benefits AI has to offer to the retail industry.

DOWNLOAD REPORT

Two-pronged approach to automated invoice processing

A problem worth solving

The average cost to process a single invoice is US$12.90. Automation delivers an average of 29% reduction in invoice processing costs, which can translate to US$300,000 per year for an organization that processes up to 10,000 invoices per month.

For decades, retailers have relied on Electronic Data Interchange (EDI) frameworks to create and send orders and to manage the supplier invoices that are generated. EDI also dramatically lowers process errors that are more likely to occur when these tasks are manually managed.

Retail-Banner

However, even the most efficiently run retailers leveraging EDIs find that nearly 30% to 40% of the invoices they must process are still managed manually. (For one large retailer that we work with, this meant processing nearly 3,000 invoices manually every day of the year.) While the EDI enables invoices to be checked automatically against predefined rules, if an invoice is not within the defined threshold, it is not released for processing. An agent from the accounts payable department is required to intervene to release or reverse the invoice. A reversal usually triggers additional actions like purchase order alterations.

Finding and framing the real problem

The challenge then is two-fold. On the one hand, find ways to ensure that the rules set in the EDI framework and the processes surrounding it are reengineered so that the EDI is able to process a greater number of ‘invoice-exceptions’. On the other hand, simultaneously automate the workflow for invoice processing being managed outside of the EDI system.

Solving the problem

  • Conduct a root-cause analysis to determine the reasons why the EDI framework is unable to process ‘invoice-exceptions’. This involves studying invoice processing and exceptions-related data over months of usage
  • Identify the top 20% of the causes contributing to 80% of the ‘invoice-exceptions’. Prioritize the task of a) reengineering processes so these can be automated through the EDI b) tweaking of decision-capability in the EDI system to be more inclusive
  • In parallel, automate the workflow for managing invoice processing outside of the EDI framework. The process typically involves manually toggling between several enterprise systems to process the task. Bring in robotic process automation that allows for a one-screen, 360 degree view of the relevant system landscape to enable agents to quickly complete the process.

The outcomes

  • Service-level guarantees for same-day processing of invoices in 80% of cases
  • Up to 45% reduction in the manual effort of invoice processing
  • Reduction in time to handle exception-invoices from seven minutes to one minute
  • Redeployment of full-time employees from managing backend invoice processing to customer-facing roles.

Use AI-led automation to amplify human potential in the enterprise

We bring robotic process automation (RPA) to automate predictable and routine operations for our clients. This enables releasing full-time employees otherwise engaged in running repetitive tasks, and creates bandwidth for them to devote themselves towards complex problem-finding and high-value innovations.