The average AR specialist spends 3-4 hours a day on follow-up that could be automated or eliminated entirely. Where to start.
When we talk to AR managers at mid-market B2B companies — the $25M to $70M revenue range where you might have two or three people managing a portfolio of 800 to 1,500 open invoices — the time-waste story is remarkably consistent. The bulk of the day goes to three things: pulling aging reports, drafting or sending reminder emails, and chasing down invoice disputes that could have been flagged earlier. Almost none of that work requires the judgment of an experienced AR professional. Almost all of it consumes them anyway.
The friction isn't just a productivity problem. It's a quality problem. When your team spends 70% of their time on routine follow-up, they have almost no bandwidth left for the accounts that actually need active attention — the customer who is genuinely disputing a line item, the one who went silent after being current for two years, the one running 45 DPD on a $40,000 invoice for the third time this year.
Where the Hours Actually Go
Before redesigning any workflow, it helps to be specific about the time allocation. In a typical AR operation at a $40M distribution company, a two-person AR team might handle roughly 1,200 open invoices at any given time. Here's a rough breakdown of where a full day goes:
- Aging report review and prioritization: 45-60 minutes. Running or downloading the aging, sorting by DPD bucket, deciding who to contact today.
- Routine outreach on current/low-risk invoices: 90-120 minutes. Sending net-30 reminders, first-notice emails, phone calls on invoices that would have resolved without contact.
- Dispute intake and routing: 30-45 minutes. Reading dispute emails, deciding whether to escalate to sales or resolve in AR, logging in ERP.
- Status updates and internal reporting: 30-60 minutes. Responding to the CFO's cash position question, updating the weekly collections status sheet.
That leaves roughly 60-90 minutes for actual high-judgment collections work: negotiating payment plans, working escalated accounts, contacting the 90+ DPD customers who are genuinely at risk. This proportion is backward. The high-judgment work should be getting the most time, not the least.
The Automation Priority Stack
Not everything in that list is equally automatable or equally safe to automate. The general rule: automate the work where the output of automation is indistinguishable from the output of human judgment — and keep humans in the loop where relationship nuance actually matters.
Start with routine dunning on low-risk invoices
The single biggest time recovery comes from automating outreach on invoices that are in the first 0-15 DPD window and have no prior delinquency history. These accounts will pay with or without a nudge. Sending them the same follow-up email your team is drafting manually each week is pure time cost with no collection lift. An automated net-30 reminder and first-notice sequence handles this volume without any human intervention — and it gets out the door consistently, which a manually-drafted queue often doesn't.
The important caveat here: this automation only makes sense when the outreach is genuinely segment-appropriate. Sending a generic "invoice is overdue" email to a 10-year customer who has never paid late in their life is noise to that customer and reflects poorly on your AR operation. Segmentation — knowing this customer's payment history — is what separates automation that helps from automation that erodes relationships.
Automate the aging pull and prioritization queue
Running an aging report and sorting it is table-stakes work that no AR professional should be doing manually every morning. An integrated AR platform should deliver a pre-sorted work queue: here are the 12 invoices that need your attention today, ranked by risk and urgency. The hour spent on aging triage disappears.
This is where predictive scoring earns its keep. Knowing which invoices are trending toward 60 DPD changes the prioritization logic entirely. Instead of triaging by days-overdue (a lagging indicator), you're triaging by predicted trajectory — and the accounts that need early intervention show up in the queue before they're technically overdue.
Build structured dispute intake, not email chains
Dispute handling is different from dunning — it requires human judgment. But the intake and routing step doesn't. A structured dispute intake process (a simple web form or in-system workflow) that captures dispute type, invoice number, and customer contact up front eliminates the back-and-forth clarification emails that eat 30 minutes per dispute. The AR specialist gets a clean ticket with the information they need to resolve it, not an ambiguous email chain to decode.
What Automation Should Not Touch
We are not saying automation replaces AR judgment. The accounts that matter most — high-value, late-chronic payers; customers showing signs of financial distress; disputes involving a sales relationship — need human voices and careful timing. Automating contact on these accounts is often worse than doing nothing. A tone-deaf dunning email sent to an account you're actively renegotiating terms with can escalate a manageable situation into a write-off.
The goal isn't to automate the AR function. The goal is to automate the parts of the AR function that don't benefit from human judgment, so that your team has time for the parts that do.
Measuring the Productivity Recovery
Two metrics are worth tracking explicitly once you've automated routine follow-up. First, the ratio of automated outreach to manual outreach — you want to see this shift over time as your team refines which segments get automated sequences versus which get personal attention. Second, and more telling, is the split of time on high-judgment versus routine work. If your AR team is spending more than 50% of their day on routine follow-up six months after implementing automation, the automation isn't working correctly — either the segments are too broad, the triggers aren't firing, or the team is manually overriding sequences they don't trust yet.
That last point is worth addressing head-on. AR specialists will continue manual work if they don't trust that automated sequences are actually going out correctly and on schedule. Adoption of automated dunning requires visibility — the team needs to see in real time which sequences fired, when, and to whom. Without that transparency, the default is manual override, and you've paid for automation that nobody uses.
A Different Starting Point
The productivity problem in AR isn't primarily a follow-up volume problem. It's a prioritization problem. When your team doesn't know which invoices are actually at risk until they're already overdue, they have no choice but to work everything — which means everything gets roughly equal attention regardless of risk. Automation of low-risk follow-up buys back hours, but it doesn't solve the underlying triage failure.
What actually changes the equation is knowing, while an invoice is still in the net-30 window, which customers are heading toward a late payment. That knowledge lets you automate the low-risk ones confidently and direct human attention to the high-risk ones early — before the situation requires the kind of aggressive escalation that damages a relationship. The productivity gain and the collection quality gain come from the same place: earlier, more accurate information about where the risk actually sits in the portfolio.
When we built the risk-scoring layer in Cashvyne, the intent wasn't to flag overdue invoices faster. It was to catch the pattern three to four weeks earlier, so that by the time an invoice crosses net-30, the AR team is already working the account with context — not reacting to an aging bucket they didn't know was building.