On-account cash is customer money that has been received and posted to a specific customer account in the AR ledger, but not yet matched to a particular open invoice. It sits as an open credit on the customer record until intent is identified and the credit is applied, refunded, or otherwise resolved.
On-account cash refers to customer payments that have been successfully posted to the correct customer subledger but have not yet been matched against a specific open invoice. The money is no longer floating in a clearing or suspense account, and the customer record clearly shows a credit balance. What is missing is the final step of the cash application process: tying that credit to the invoice or invoices it is intended to settle.
This is where confusion often creeps in. Many AR teams use on-account cash and unapplied cash interchangeably, but the distinction matters for both ledger hygiene and reporting accuracy. Unapplied cash is money that has hit the bank and the cash account, but has not been posted to any customer at all. It sits in suspense waiting for someone to identify the payer. On-account cash has cleared that first hurdle. The payer is known, the customer balance has been updated, and the open issue is purely an invoice-matching problem.
Treating the two as the same thing leads to inaccurate aging reports, misleading collector worklists, and weaker control over customer credit balances. Keeping them separate gives finance leaders a much sharper view of where friction is actually happening in the order-to-cash cycle.
On-account credits build up for a handful of recurring reasons, most of which are operational rather than malicious. Understanding the root cause is the first step toward resolving the balance correctly.
On-account cash quietly distorts almost every receivables metric finance leaders care about. The most visible effect is on Days Sales Outstanding. Because the original invoice remains open in the ledger, DSO calculations continue to treat it as uncollected, even though the cash is already in the bank. A company with a healthy collections operation can look slow simply because cash app has not closed the loop.
The effects do not stop at DSO. Customer statements become confusing, showing both an open invoice and a credit balance, which often prompts the customer to ask why they are being chased for money they already paid. Aging buckets get noisier, collectors waste time chasing the wrong accounts, and credit teams may freeze orders for customers who are in fact paid up. In multi-entity or multi-currency environments, on-account balances can also cause intercompany reconciliation breaks and revaluation noise at period end.
A disciplined resolution workflow stops on-account cash from quietly building up over time. The basic shape of the process is the same across most AR teams.
One of the most overlooked risks with on-account cash is escheatment, also known as unclaimed property reporting. In many jurisdictions, including most US states and a growing number of European countries, customer credits that sit unresolved for a defined dormancy period must be reported and remitted to the relevant authority. Writing aged credits off as miscellaneous income to clean up the ledger is not a safe shortcut. It can create audit findings, tax exposure, and in some jurisdictions, financial penalties for non-compliance with unclaimed property law.
The safer posture is to treat on-account cash as a controlled balance with documented dormancy thresholds, customer outreach steps, and a defensible escheatment process for anything that genuinely cannot be returned or applied. Internal audit and external auditors increasingly look at how this is managed, particularly in companies with large customer bases or recurring billing models.
Modern, AI-native cash application platforms attack on-account cash on two fronts. First, they reduce how much of it gets created in the first place. Agentic systems read incoming payments even when remittance is missing or unstructured, then predict the most likely invoice match using customer payment history, amount patterns, timing relative to invoice dates, and historical behaviour. When confidence is high enough, they auto-apply directly to invoices instead of dropping the cash on-account.
Second, they actively manage the residual on-account balance. Agentic workflows continuously monitor customer credits, flag balances that are aging past defined thresholds, and route them to the right owner with a recommended action: apply against the next invoice, request remittance, refund, or escalate. Combined with clean ledger discipline, this approach keeps on-account cash low, keeps DSO honest, and keeps escheatment risk under control without burying the team in manual review work.
On-account cash has been posted to a specific customer record but has not yet been matched to a particular invoice. Unapplied cash has been received into the bank or a clearing account but has not yet been posted to any customer at all. The distinction matters because each problem has a different root cause and resolution path.
DSO is calculated based on open invoices, not on customer credit balances. When cash sits on-account instead of being applied, the underlying invoice stays open in the ledger and continues to count as outstanding receivables, even though the money is already in the bank. This makes collections look slower than it actually is.
At minimum monthly, with tighter weekly review for large customers and high-value credits. Anything aging past 90 days should be on an exception report, and balances older than 180 days need a documented resolution plan to manage escheatment exposure and customer-statement accuracy.
Generally no. In most jurisdictions, unresolved customer credits are subject to unclaimed property or escheatment regulations after a defined dormancy period. Writing them off as miscellaneous income can trigger audit findings, tax issues, and regulatory penalties. The correct path is documented customer outreach, refund where appropriate, and formal escheatment for genuinely dormant balances.
When the customer has a clear pattern of prepaying, when a future invoice is already scheduled, or when the customer explicitly confirms that the credit should be used against upcoming charges. Guessing without evidence often creates disputes, so the safer default is to confirm intent before applying credits to invoices that were not part of the original payment.
AI-native systems use customer payment history, amount matching, and timing signals to predict which invoice an incoming payment is meant to settle, even when remittance advice is missing or unstructured. High-confidence matches are auto-applied directly to invoices, so the cash never lands on-account in the first place. Residual credits are continuously monitored and routed for action before they age into a compliance problem.
