Blind Spots in Enterprise Invoice Processing: Two-Fifths of Invoices Have Issues, Involving Up to $549 Billion
A study based on data from 27.2 million invoice processing records shows that 40% (or two-fifths) of invoices globally arrive without purchase orders, resulting in an annual blind spot of approximately $549 billion in unmanaged spending. Agentic AI solutions like Basware's InvoiceAI are expected to help enterprises automate the processing of non-PO invoices, reduce manual intervention, and improve financial efficiency.
Charlotte, North Carolina — Latest data shows that two-fifths (40%) of invoices at global companies arrive with issues, creating an annual blind spot of approximately $549 billion in unmanaged spending that not only drains finance team resources but also exposes companies to risk.
Large organizations process thousands of non-PO invoices each year. When invoices arrive without a purchase order, companies have less control over spending due to the lack of a pre-approval process, often leading to delayed supplier payments.
According to Basware's analysis of 272 million invoice processing records last year, 40% (two-fifths) of invoices at global organizations arrive without a purchase order. This means companies face an annual "blind spot" of approximately $549 billion in unmanaged spending.
Agentic AI offers a solution for finance teams, providing the visibility and control needed to manage hidden spending.
How does Agentic AI work?
Agentic AI can transform invoice processing for finance teams—autonomously reading, validating, and matching PO and non-PO invoices, reducing process bottlenecks without extensive manual touchpoints.
Without Agentic AI:In non-AI and manual systems, an invoice lacking a PO number gets set aside because it requires manual review, coding, routing, and approval.
With Agentic AI:When an invoice without a PO number arrives, AI reads it, understands its purpose, suggests the correct coding based on similar historical invoices, and routes it to the appropriate approver.
Agentic AI acts like a skilled accounts payable (AP) clerk, capable of thinking, deciding, and solving problems—rather than merely a tool that executes instructions.
How Agentic AI saves hundreds of hours
Accounts payable departments spend 10% of their time processing non-PO invoices, equivalent to over 200 hours per employee per year, leaving the CFO's office under-resourced without knowing it.
By embedding Agentic AI into financial operations, companies can gain visibility and accuracy on every invoice, improving compliance and on-time payment rates.
InvoiceAI enables touchless processing
To address the $549 billion blind spot, Basware has launched InvoiceAI, a suite of 5 AI-enhanced features designed to optimize the invoice lifecycle management (from delivery to approval and payment):
- SmartCoding – Automatically assigns the correct coding to invoices, eliminating manual work for finance teams.
- SmartWorkflow – Routes invoices to the correct approver without requiring human input.
- SmartPDF – Converts paper or PDF invoices into instantly processable digital data.
- AP Business Agent – Provides real-time guidance to employees while processing invoices, suggesting the next best action.
- AP Data Agent – Allows employees to get simple answers about invoices and generate reports through natural language processing.
InvoiceAI helps CFOs and AP teams improve efficiency by increasing the number of invoices processed without issues. For invoices that do require human intervention, InvoiceAI provides clear recommendations in a centralized hub for handling the two-fifths of non-PO invoices.
Perttu Nihti, Chief Product and Technology Officer at Basware, commented:
"Finance teams face immense pressure to reduce costs while improving operations. The difference is this: traditional AI is like a calculator—you need to input the equation to get an answer. Agentic AI doesn't wait for instructions. It sees a non-PO invoice, understands what needs to be done, and recommends a course of action. Non-PO invoices are a long-standing challenge, but with Agentic AI, the 200 hours spent processing them each year can be eliminated."
Real-world applications of Agentic AI
Global cable manufacturer Nexans is an example of a company leveraging AI to improve the operational efficiency of its finance functions.
By implementing Basware's AI enhancements (including SmartPDF), Nexans reduced invoice processing time from days to hours, while the AP team gained full visibility into the incoming invoice flow thanks to its tracking capabilities.
Cecile Barrere, AP Director at Nexans, said: "Basware SmartPDF gives us better visibility and traceability of invoices and reduces the risk of errors and fraud. In just 7 weeks, we achieved 92% reliability in automatic identification and processing of invoice data with the help of AI and machine learning."
