Late payment predictor

The late payment predictor anticipates the likelihood of an invoice being paid late, and is generated using a machine learning model.

The late payment predictor is shown for all customers who have at least one paid invoice for your organisation. The score is generated for invoices that are before due, do not have a payment plan configured, or where the expected payment date is prior to the invoice due date.


The late payment predictor categorises risk of late payment as low, medium or high. Hover over the rating to see the score displayed as a percentage. 





The invoice has a low likelihood of being paid late (0%-50%)


The invoice has a moderate likelihood of being paid late (50%-75%)


The invoice has a high likelihood of being paid late (75%-100%)


How can I use the late payment predictor to improve my credit control processes?

Use the late payment predictor to:

  • Understand your current risk (how many invoices may be paid late)
  • Prioritise your credit control efforts
  • Make the most impact by targeting the highest risk invoices


How is the late payment predictor calculated?

The late payment predictor takes into consideration your customer's previous payment behaviour and predicts their future payment behaviour on their due invoices.

The calculation takes into consideration the following:

  • Invoice due date
  • Invoice value
  • Recent and historical payment behaviour and invoice details
  • Number of invoices paid late most recently and historically

NB: The late payment predictor will only be available for CSV integrated organisations if you upload the payments file (optional). This tool is not available for AccountsIQ or Zohobooks integrations.