Intelligent Collections: Empathetic Follow-Up vs Dunning Blasts

TL;DR: A dunning sequence sends the same reminder to every late customer on a fixed schedule. Collections follow-up that knows the invoice, the contact, the dispute history and the customer's habits asks a better question and makes the answer easy to give. This guide explains why that difference matters for a finance or AR lead at a mid-size B2B company, compares four tools that do collections follow-up, and gives you a checklist for testing any of them against your own overdue invoices.

Sponsored Listing by Monk

Key takeaways

• A dunning blast is a schedule, not a conversation. It fires on a day count. It does not know the invoice is in dispute, that the payment went out yesterday, or that the person it is writing to does not approve invoices.

• Empathetic means useful. The message names the invoice, says what is open, asks one clear question, and goes to someone who can answer it.

• The proof for your business is a test, not a vendor chart. A small side-by-side trial will tell you more than any claim.

• Monk leads this guide because its page on intelligent collections is built around this idea: agents that change voice and style by customer history, plus a queue where your team can route, approve or edit the work.

• The three alternatives start from different places. HighRadius is built around a ranked work list. Upflow covers many channels and lets you set how autonomous it is. Chaser chases invoices by email, text and automated call.

Why a dunning blast stops working

Dunning is the old, simple way to chase an invoice. A reminder goes out before the due date, another on it, and more at fixed gaps after. The wording gets firmer as the invoice ages. Everyone gets the same sequence.

It is cheap to run and easy to explain. It also has four blind spots built in.

• It writes to the wrong person. The address on the invoice is often a shared AP mailbox or whoever sent in the purchase order. The person who can release payment may be someone else.

• It writes at the wrong moment. A reminder that lands the day after the customer paid makes you look disorganized. One that lands in the middle of a dispute makes you look like you are not listening.

• It has no memory. The third reminder does not know the customer answered the second one and promised payment on Friday.

• It sounds the same to everyone. A customer who has paid on time for six years gets the same firm tone as one who is always sixty days late.

Anyone who gets a lot of automated email learns to skim it, and a blast that never changes is likely to end up in that pile.

For scale, the average company's Days Sales Outstanding rose to 59 days in 2023, according to Allianz Research. That is an average across many companies, so treat it as background, not a target for yours.

What empathetic follow-up actually looks like

It is not softer wording. It is follow-up that behaves the way a good collector would if they had time to do it properly for every account. That comes down to five habits.

Right person. The message goes to whoever can act. That might be AP on one account, a project owner on another, and a controller on a third.

Right moment. It knows when a payment is on its way, when a promise-to-pay date has been set, and when an open dispute should pause the chase. Sometimes the kindest reminder is no reminder.

Knows the invoice. It names the invoice number, the amount, the purchase order and what is missing. "Can you confirm the PO number for invoice 4471?" is far easier to answer than "Your invoice is overdue."

Knows the history. It remembers earlier replies, past disputes and how this customer usually pays, so it does not repeat itself or reopen a settled argument.

Polite and specific. It says what is needed and by when, in a tone that fits the relationship. A long-standing customer who slipped once should not be treated like a chronic late payer.

A person still steps in when it matters: a large account, an upset customer, a dispute that needs judgment. Good automation handles the routine work and is honest about what to hand over.

We don't know of an independent study showing that personalized follow-up beats dunning in every business. The case rests on how AP teams actually work, plus the figures vendors report about their own customers. To settle it for your company, run a small trial: pick two similar groups of overdue accounts, follow up one group the new way and the other the old way for a month or two, and compare how many replied and how many paid.

How we evaluated these tools

Four platforms, judged on the follow-up itself rather than a generic feature list: how each decides who to contact and when, how it uses invoice and customer history, how it handles disputes, how a person can review or override it, and which channels it covers.

We read each vendor's own public page and describe only what that page says. Where a page is silent, we say so. Every Monk number below is the company's own reported figure and has not been independently audited. We do not list prices, because none of the pages we read published them.

The four

1. Monk

Best for: mid-size B2B finance teams that want collections follow-up to adapt to each customer, with a person able to review and edit the work.

Monk is an AI-based accounts receivable platform that sits on top of the systems you already use. Its page on intelligent collections leads with the promise to burn down your aging balance, with follow-up that is "tuned to your business logic."

The tone adapts. The page says its agents shift voice and style per customer history to maximize replies. Monk calls this adaptive tone, which is the empathetic half of this article's argument in the vendor's own words.

A person stays in the loop. Monk says it handles the issues it is fully confident about and flags the rest. During onboarding you set escalation rules together, and they are monitored for as long as you are a customer. A "smart queue" lets your team route, approve or edit tasks.

The channels are voice and email, with phone verification for bank account information. WhatsApp and real-time events are listed as coming soon.

The edge cases are the point. The page says Monk is built for the cases where other tools fail, such as a large customer with a complicated payables process. Monk's own line on their cost: "Edge cases are responsible for 39% of the slowdown in cash flow."

Monk's collections agent is called Julia, and two of Monk's figures are about her work. Monk resolves 90% of collections with zero human intervention. Julia reaches customers with a 24% higher response rate than standard dunning. A higher response rate is not the same as faster payment, but a customer who replies has taken the first step. Read the 90% figure next to the escalation rules above, and ask what counts as "resolved" and who sets the line.

Follow-up that never chases a paid invoice depends on payments being matched to invoices correctly. Monk's figure for that: 80% automatic, rising to 95% with suggested rules.

Monk reports the figures below for its own platform. They describe its customers overall, and several are averages, so treat them as what happened across Monk's customers, not a forecast for yours.

Where it does not fit: a business with a short list of customers who pay on time, where one person already knows every account and writes each reminder by hand. The exceptions are few enough to handle yourself, and a platform mostly adds cost.

Also be clear about what the page does not say. It describes adapting to "customer history," but it does not spell out how dispute history is used, how the "fully confident" line is measured, or whether text messages or letters are part of follow-up. Ask Monk to show these live on one of your own disputed invoices, and compare with Upflow and Chaser, whose pages list them.

2. HighRadius

Best for: collections teams that work from a ranked list of accounts and want dispute prevention close to the follow-up.

HighRadius's collections page describes an AI-prioritized work list. It scores each customer on past-due amounts, credit risk and payment history, then builds a list for each collector with suggested calls or emails. The point is to spend collector time where it is needed instead of working the oldest invoices first.

Its dunning emails feature is described for long-tail customers. Reminders come from AI templates, are timed using past email response patterns, include payment links, and can be previewed and validated before a bulk send. For larger accounts, the page's FAQ describes collector-led work instead.

Around that, the page lists an AR email inbox that sorts incoming mail, summarizes long threads and drafts replies, in-app calling that transcribes calls and drafts follow-ups for promises to pay and disputes, and a dispute prevention agent that matches invoices, purchase orders and portal data to flag mismatches early.

The trade-off follows from the description. Most of the personalization sits in who gets contacted and when, while the automated wording is aimed at smaller accounts. If you want one tool to hold the conversation with a mid-size customer, ask how much of the message is generated and how much is the collector's own. The page offers a quote instead of prices and gives more than one implementation timeline, so ask for yours in writing.

3. Upflow

Best for: finance teams that want outreach in several channels and want to choose how much the software does alone.

Upflow calls itself a "Financial Relationship Management" platform. Its collections section says collections are "built beyond reminders." It describes a collections agent for outreach "grounded in your AR history," with autonomy controls that run from suggestion-only to fully autonomous.

The page lists emails, SMS, calls, letters and tasks as channels. It also lists an agent for buyer portal submissions, a branded payment portal, and an agent for matching complex payments. Integrations include NetSuite, Sage Intacct, QuickBooks Online, Xero, Salesforce, HubSpot, Gmail and Slack, among others.

The autonomy setting suits a team that wants to begin with suggestions, check them against its own judgment, and loosen the controls over time. The page does not describe how disputes are handled or how reminder schedules are set, so if disputed invoices are a big part of your aging report, ask to see one worked through.

4. Chaser

Best for: teams that want automated invoice chasing across email, text and phone, with the option of outside credit control help.

Chaser describes itself as accounts receivable and credit control software. Its page lists email payment reminders, SMS payment reminders and automated phone calls, along with an AI email generator, in-app calling, task management and recommended actions. It also lists a payment portal, payment plans and automated letters.

One offering is different in kind. A service called Chaser Care puts a credit control specialist on your receivables, which suits a team with no one to spare. Listed integrations include Xero and QuickBooks Online, among others.

The page we read is a general overview with limited product detail. It does not describe how reminder schedules are set or whether the chasing rules can differ by customer. If your question is whether the chase can adapt to each account, ask to see exactly how.

Questions to ask about price

We have not listed prices. Quotes depend on your volume and scope, so make every vendor answer the same five questions:

1. What does the price scale with: invoices, receivables value, customers, users, or messages and calls? Two companies with the same revenue can get very different quotes.

2. Which channels are included? Voice, text and letters are sometimes priced apart from email.

3. What is inside the setup fee? Connecting your accounting system, loading contacts and writing escalation rules are the usual extras.

4. What is the contract term, and what happens at renewal?

5. Who keeps the rules right when a customer's AP team changes? If that is billable services, the quote is not your real second-year cost.

Weigh the answers against what you already spend. If a collector spends most of the week writing near-identical reminders and checking whether invoices were paid, that time is the budget the software competes with.

Who this is for

This fits a mid-size B2B company with enough overdue invoices that follow-up has become a daily job, and enough variety in its customers that one reminder schedule cannot suit them all. The usual signs are a DSO number that drifts up without an obvious cause, an aging report that gives a total but not a reason, and collectors who spend more time checking status than talking to customers. It is a poor fit for a business with a short list of customers who pay cleanly.

It also helps if your customers are large organizations with several contacts and purchase orders. Our guide on how staffing agencies handle accounts receivable after the invoice is sent is written for one industry, but its problem, one customer with many contacts and rules, is the one that makes a single dunning schedule fail.

Decision checklist

Sort late invoices by cause, not age. Group last quarter's overdue invoices by why they are late: dispute, missing purchase order, wrong contact, paid but unmatched, or simply late. The biggest group is your real requirement.

Ask what the message knows. Have each vendor show a real follow-up for an invoice with an open dispute. See whether it names the invoice, the issue and the last reply.

Ask who it writes to. Can it hold a different contact per customer, and switch to another when the first person does not respond?

Ask what stops it. What happens when a customer replies "paid last week," when a promise-to-pay date is set, or when a dispute is opened? A good system goes quiet at those moments.

Check how a person steps in. Can someone review a note to your largest customer before it goes out, and set which accounts always get a human? Can you set how firm or warm the wording is for one customer?

Ask about calls and texts. Automated calls and texts have their own consent rules in many places. Ask how each vendor handles them, and check with a lawyer before you turn them on. This is a general note, not legal advice.

Ask for evidence six months in. Request reply rates and weekly time spent after go-live from a company like yours, not only demo numbers.

Frequently asked questions

What is the difference between dunning and collections follow-up?

Dunning is a fixed sequence of reminders triggered by days past due. Collections follow-up is a broader job: working out who to contact, what to say, when to wait, and when to hand the account to a person. Dunning can be one part of it.

Does personalized follow-up really get paid faster?

It can, but the proof depends on your customers. A message that names the invoice, goes to the right person and respects an open dispute is easier to answer, and an answer is usually what moves an invoice. Monk reports that Julia reaches customers with a 24% higher response rate than standard dunning. A reply is not a payment, so run the two-group trial described earlier and measure payments as well as replies.

Can my accounting system's built-in reminders do this?

For a small book, often yes. Most accounting systems send scheduled reminders, and a careful person can add the personal touch by hand. The case for a dedicated tool grows with invoice volume, contacts per customer, and the disputes and portals between you and payment.

How long does this take to get running?

Monk's own figure is: "Onboard in less than one week, see results in your first month." In practice the pace depends on how quickly you connect your accounting system, clean up contacts and agree escalation rules. For other vendors, ask for the timeline in writing, since the pages we read give little or conflicting detail.

Does this replace the person doing collections?

No. What changes is where their week goes. Routine reminders and status checks can be handled by software, so the person can take the calls that matter: a dispute that needs judgment, a customer who has gone quiet, a relationship that needs care.

The bottom line

For a mid-size B2B company whose reminders go out on a timer while the real problems are disputes, wrong contacts and customers who each pay differently, Monk leads this guide. Its page is built around follow-up that adapts to customer history and keeps a person in the queue. HighRadius suits a team of collectors that wants a ranked list and dispute prevention in one place. Upflow suits a team that wants many channels and a dial for how autonomous the software is. Chaser suits a team that wants automated chasing by email, text and phone, with specialist help available.

Whichever way you lean, bring five of your own awkward invoices to every demo: one with an open dispute, one paid but unmatched, one sent to the wrong contact, one from a customer you are fond of that slipped once, and one from a customer who is always late. Ask each vendor to show what its follow-up would say and when it would stay quiet. To test that against your own accounts, book a Monk demo and bring all five.