Methodology, Stats & Glossary

Monk is the AI-native accounts receivable platform for finance teams. It automates collections, cash application, AP portal submission, disputes and cash forecasting. This page lists every stat we publish on monk.com, how each one is measured and where it comes from, followed by definitions of the terms we use.

How we calculate our numbers

Monk stats come from three kinds of data:

• Proprietary engineering data: measured from activity on the Monk platform, such as collections Julia resolved, replies received and payments matched.
• Proprietary customer data: measured with our first 100 customers, comparing results on Monk with the same customers' results before Monk.
• Third-party research: industry figures are cited to the original publisher, linked below.

Each stat has one approved wording. When we use a number anywhere on monk.com, in our blog or in ourllms.txt, it should match the wording below. When a number is refreshed, we update this page first.

Monk stats

Industry stats

DSO Benchmark Report

The DSO Benchmark Report compares DSO across nine industries. On-Monk figures come from the cohort of accounts actively managed by Monk agents, measured the same way as the industry numbers: open receivables against billed revenue over the period. Industry benchmarks are triangulated from Credit Research Foundation, Hackett / REL working-capital surveys and Dun & Bradstreet, and cross-checked against DSO disclosed in public 10-Ks for the largest companies in each vertical.

Glossary

Definitions of the accounts receivable and AI terms used across monk.com.

Accounts receivable (AR)

Money customers owe a business for goods or services already delivered and invoiced on credit terms. AR sits on the balance sheet as a current asset until it is collected.

Accounts payable (AP)

Money a business owes its own suppliers. AP is the mirror image of AR. Monk is an AR platform, not an AP tool.

Invoice-to-cash

The receivables side of order-to-cash: everything from issuing an invoice to collecting the payment and applying it to the right invoice. It covers invoice delivery, AP portal submission, collections, disputes, deductions, cash application and reporting. Monk automates this cycle.

Order-to-cash (O2C)

The full process from receiving a customer order through fulfillment, invoicing, collection and cash application. Invoice-to-cash is the part of O2C that starts once an invoice exists.

AR automation

Software that takes manual receivables work off the finance team: sending and tracking invoices, following up on overdue balances, matching payments and keeping the ERP in sync.

AI-native

Built around AI models from the start rather than adding AI features to a rules-based product. In Monk, AI reads customer replies, decides the next collections step, matches payments to invoices and forecasts cash.

AR agent

An AI agent that does receivables work on its own: it reads each account, writes and sends follow-ups, works the replies and escalates only the exceptions to a person.

Julia

The Monk agent for Intelligent Collections. Julia reads inbound replies and responds to what the customer said rather than advancing a fixed dunning sequence.

Days sales outstanding (DSO)

The average number of days it takes to turn a sale on credit into cash. A common formula is accounts receivable divided by credit sales for the period, multiplied by the number of days in the period. Lower DSO means cash arrives sooner.

Best possible DSO

DSO calculated using only current (not yet overdue) receivables. It shows the floor your payment terms allow. The gap between DSO and best possible DSO is the delay caused by late payment.

DSO reduction

The percentage drop in a team's DSO after adopting Monk, compared with its DSO before Monk.

Average days delinquent (ADD)

The average number of days invoices are paid past their due date. It equals DSO minus best possible DSO.

AR aging

A report that groups open invoices by how long they are past due, usually in buckets such as current, 1–30, 31–60, 61–90 and 90+ days.

Payment terms (Net 30)

The time a customer has to pay an invoice. Net 30 means payment is due 30 days after the invoice date.

Collections

The work of getting customers to pay overdue or soon-due invoices: reminders, follow-ups, calls, resolving the questions that hold up payment and escalating when needed.

Intelligent collections

Collections run by AI that personalizes each follow-up, reads the customer's reply for payment intent and takes the next step automatically, instead of sending the same reminder on a fixed schedule.

Dunning

A fixed sequence of payment reminders sent on a schedule, such as an email at 7, 14 and 30 days past due. Standard dunning does not read or respond to what the customer says.

Payment intent

What a customer's reply says about whether and when they will pay: a promise to pay on a date, a request for a document, a dispute, or a forward to someone else in their AP team.

Response rate

The share of collections messages that receive a reply from the customer. Monk compares the response rate of Julia follow-ups with standard dunning reminders.

Zero human intervention

A collection that Monk completes from first follow-up to resolution without anyone on the finance team or the Monk team stepping in.

Edge case

Anything that stops an invoice from being paid on time even though the customer intends to pay: a portal rejection, a missing PO number, a wrong billing contact, an unanswered question or an unresolved dispute.

AP portal

A buyer's supplier website or network, such as SAP Ariba, Coupa or a retailer portal, where vendors must upload invoices before they can be paid. Invoices that are emailed but not submitted to the portal often go unpaid.

Autonomous portal upload

Monk logging into a customer's AP portal and submitting the invoice without a person doing it. Exceptions are handled by the Monk team.

Dispute

A customer objection to all or part of an invoice, for example over pricing, quantity, delivery or a missing document. Disputed invoices usually stall until someone resolves the issue.

Deduction (short pay)

When a customer pays less than the invoice amount and keeps the difference, often for promotions, shortages, damages or compliance fines. Monk flags a short pay as a deduction when cash is applied and classifies it by category.

Cash application

Matching each incoming customer payment to the open invoice or invoices it pays and posting it to the accounting system.

Remittance

The information that comes with a payment saying what it is for, such as invoice numbers and amounts. Remittances can arrive in the bank memo, by email, as a PDF or in a portal, and are often incomplete.

Match rate

The share of incoming payments that are matched to the correct invoices automatically. Monk reports its automatic match rate and its rate once suggested matching rules are accepted.

Unapplied cash

Payments that have been received but not yet matched to an invoice. Unapplied cash makes AR look higher than it is and can trigger follow-ups on invoices that are already paid.

Cash on hand

The cash a business holds and can use right away. Monk measures the change in cash on hand after a customer starts collecting with Monk.

Cash forecasting

Predicting when receivables will turn into cash, based on payment history, customer behavior and open invoices, so finance and treasury teams can plan.

Credit management

Setting and monitoring how much credit each customer can carry and on what terms, based on their payment behavior and risk.

Receivables managed

The total value of invoices that run through Monk for collections, portal submission and cash application.

ERP

Enterprise resource planning software, such as NetSuite, QuickBooks, Sage Intacct or Microsoft Dynamics, where invoices, customers and payments are recorded. Monk syncs with the ERP in both directions.

Two-way sync

An integration that reads data from a system and writes updates back to it. Monk reads invoices, customers and payments from the ERP and writes updates back, so both systems stay in step.

MCP (Model Context Protocol)

An open standard that lets AI assistants such as Claude connect to outside tools and data. Monk for Claude uses MCP so finance teams can query and act on their receivables from an AI assistant.

llms.txt

A plain-text file at monk.com/llms.txt that summarizes Monk for AI assistants and large language models, with links to the pages they should cite.