A division of an industrial Fortune 50 company wanted to improve its working capital and reduce outstanding invoices, which totaled over $800 million. The company’s finance team used the Maana Knowledge Platform to analyze over five years of invoices to uncover hidden patterns and correlations, as well as identify specific recommendations that would reduce days sales outstanding.

As shown in the figure below, the Maana Knowledge Platform crawled and mined data related to over one million invoices across a wide range of business silos, including historical data (such as open, closed and disputed invoices; collector logs and customer loyalty information) and external data (such as customer credit ratings, the price of oil and interest rates). Maana also used the time value of money as part of its analysis.

After creating and evaluating several data models using the Maana Knowledge Platform, the company was able to accurately predict the likelihood of late payments before invoices were actually due, as well as identify the root causes of late payments, which included:

  • Weekend due dates: The analysis showed that any invoices with due dates of Saturday and Sunday were always late.
  • Lack of familiarity with invoices for first-time customers: The majority of first-time invoices were usually late by 90 days because the customer did not understand their invoice.
  • Customer satisfaction: Many late invoices were attributed to service issues that had not been addressed and required customer service follow-up, not a collection call.

Business Benefits

  • Improved A/R collections by 65% over the prior year.
  • Recommended four collection call queues to be embedded into the GETPAID system:
    • First-time customers
    • Customers with service issues
    • Institutional customers
    • Customers with late invoices for other reasons

Using the insights gained through the Maana Knowledge Platform regarding the root causes of late payments, the company created a customized call list for each collection agent; this call list was incorporated into the organization’s AvantGard GETPAID collection system.

As the Maana Knowledge Graph learns and adapts over time, it provides ongoing, data-driven recommendations regarding which customers should be called and when. For example, at one point, it recommended that accounts payable call all new customers ten days prior to their invoice due dates to ensure they understand their invoice and get questions answered.

After just 30 days of operationalizing recommendations like this one into GETPAID, the Maana Knowledge Graph and machine learning algorithm continues to learn, adjust and fine-tune insights and recommendations based on daily data input on open and closed invoices, collector actions and stock market changes. For example, the platform recently identified four groups of customers as consistently late payers, as well as specific strategies that the company can take to mitigate future late payments:

  • First-time customers: The Maana Knowledge Platform recommended that finance make a courtesy call to these customers to ensure they understand their invoice and can ask questions at least 10 days prior to invoice due dates.
  • Customers with unresolved service issues: The Maana Knowledge Platform recommended that the finance department call customer service to ensure open cases are resolved, as most customers with unresolved issues will not pay.
  • Institutional Customers: These customers had contracts with longer net payment terms, so accelerating collections would require re-negotiating the contract terms.
  • Other: Accounts receivable clerks should call all other customers with late payments for other reasons.

By operationalizing all of these recommendations, the company improved A/R collections by 65% over the prior year, which increased working capital by $520M per year.

Digital Transformation Initiatives



Minimizing Insurance Policy Risk

Talk To A
Maana Specialist

Ask our product experts how
Maana can help increase your
asset profitability.

Contact Us