Artificial Intelligence & Automation
Predictive models, generative AI and agents, trained on your existing data and delivered in the dashboards your teams already use.
Learn moreFinance · Finance department of an industrial group
Predicting the collection date of every customer invoice
Replacing theoretical due dates with the actual payment behaviour of each customer account, and delivering the forecast in the report the finance team already opens.
Data scope
| Source | Finance datamart — customer accounts |
|---|---|
| Volume | Consolidation of several group ERPs |
| Depth | Invoices since 2017 |
| Delivery format | BI dashboard by customer account |
Context
This industrial group’s cash forecast is built on the contractual due dates of customer invoices. Yet customers pay according to their own behaviour, and the receivables ledger is consolidated from several group ERPs in the Finance datamart.
Problem
The theoretical date is rarely the actual date, and the gap is discovered at month end. The cash forecast therefore rested on due dates that customers do not keep.
Objectives
Predict the payment date invoice by invoice, derive the expected incoming cash flow, and objectify the payment behaviour of each customer account.
SOFALI’s intervention
The same chain as for a supply chain predictive project, applied to the receivables ledger.
Separation and filtering
The fact table was split into settled invoices and open invoices. Partial payments, credit notes and unallocated entries were excluded; the scope was limited to customers with a sufficiently dense history.
Learning the behaviour
The model is trained on settled invoices: for each account, it learns the usual gap between due date and actual payment, then applies it to the invoices still open.
Delivery to management control
For each customer account and each month, the amount collected, the amount due and the predicted amount are compared on a single chart, with the group’s organisational hierarchy and a factoring filter.
Going live
A “predicted payment” column was added to the existing customer accounts report, together with a dedicated tab for the prediction values. The forecast arrives in the tool the finance team already opens, with no new reporting to adopt.
Data and technology
A model trained on settled invoices to learn the payment behaviour of each account; SQL to separate and filter the invoices; a BI dashboard by customer account for delivery.
Results
Model delivered and integrated into the customer accounts report, on a receivables ledger consolidated from several ERPs. No accuracy figure published.
Gains identified
- Steer incoming cash flow on a forecast rather than a theoretical due date.
- Reduce bank charges: transfers, foreign exchange operations.
- Reduce overdraft charges and reliance on short-term credit lines.
- Objectify the payment behaviour of each customer.
- Reduce the average settlement time.
- Negotiate payment terms on the basis of the actual history.
Documented limitations
We deliver a model's limitations together with its results.
- Extend the model to customers with a low invoice volume.
- Integrate partial payments through the date of the last payment.
- Set up cross-validation.
Related expertise
BI & Data Analytics
Data warehousing, modelling and reporting: a 360° view of your business and your customers, built on solid foundations.
Learn moreQuantified results correspond to achievements measured on the scopes and datasets described; they do not constitute a performance commitment on any other scope. Offers flagged as "to be scoped" describe a proposed approach, not an existing achievement.
A similar project? Let's talk.
A first conversation to frame your needs, with no commitment, in Paris or Casablanca.