The dunning process — the structured communication with customers about overdue invoices — is one of the most critical yet underestimated parts of the accounts receivable (AR) function. When executed well, it accelerates cash collection, reduces Days Sales Outstanding (DSO), and maintains strong customer relationships. When managed poorly, it leads to missed payments, damaged trust, and higher bad debt.
Here are five best practices modern finance teams use to make their dunning workflows more effective, consistent, and customer-centric.
1. Personalise communication
Generic reminders are easy to ignore. Customers recognise when they're receiving an automated template instead of a genuine message. Personalised communication — referencing specific invoices, payment history, and prior interactions — creates accountability and drives faster responses.
AI agents can now personalise at scale. They continue ongoing threads, adapt tone based on behaviour, and sound like a member of your AR team across thousands of customers.
2. Find the right contact information
Many overdue invoices are caused by messages sent to the wrong inbox. Contacts change roles, AP departments reorganise, and shared addresses get ignored. Validating and maintaining accurate contact data is key to successful collections.
AI agents can automatically detect bounce-backs, identify new contact details, and suggest updates when emails go unanswered. This ensures your reminders actually reach someone who can take action — not a dead inbox.
3. Track and act on promise-to-pay dates
When a customer commits to paying next week, that's valuable intent data. Too often, it's lost in email threads or forgotten in spreadsheets. AI agents can capture these promise-to-pay dates automatically, pause reminders until the commitment date, verify payment once it's due, and resume outreach if necessary.
This keeps communication professional and efficient while avoiding unnecessary follow-ups with customers who are genuinely planning to pay.
4. Involve sales and account management
Collections are not just a finance task — they're part of the overall customer relationship. When key accounts become overdue, sales and account managers can provide context that AR may not have: ongoing negotiations, open renewal discussions, or service issues that might explain delayed payment.
Involving customer-facing teams early prevents escalation from damaging relationships. AI agents can flag high-value or at-risk accounts automatically, notifying the right people in sales or customer success to intervene before situations escalate.
5. Escalate intelligently — at the customer level, not the invoice level
Escalation should reflect the broader customer relationship, not a single overdue invoice. Treating every invoice in isolation can create awkward situations — like aggressively chasing a customer over a 120-day-old €120 invoice when they've already paid five others on time.
Instead, look at the customer's total exposure: number of outstanding invoices, total overdue balance, and payment history. Escalation should happen at the customer level, taking all open items into account.
AI agents can automatically calculate the total overdue amount and determine when to escalate, ensuring follow-ups are appropriate, prioritised, and aligned with the customer's overall payment behaviour.
The future of dunning
The best finance teams are moving from static, rule-based reminders to dynamic, conversational workflows powered by AI. AI agents now handle thousands of customers simultaneously, personalise outreach, find the right contacts, track promises to pay, involve account owners when needed, and escalate intelligently at the customer level.
The result: faster collections, lower DSO, fewer disputes, and better customer experiences — without adding headcount or outsourcing collections.
Want to see how AI agents can transform your dunning workflows? Book a demo with Paraglide AI.