Invoice Exception Handling Guide

Invoice exception handling is the accounts payable process of resolving invoices that cannot be paid straight through because something is wrong, missing, or does not match the purchase order or the goods receipt. Typical exceptions are a price or quantity mismatch, a missing PO, a duplicate, an unapproved vendor, or a math error. Handling them means routing each one to the person who can resolve it, fixing the root cause, and paying the invoice once it clears. A small share of invoices create most of the AP workload, and that share is where late payments and missed discounts come from. Last updated July 2026.

This guide explains what an invoice exception is, what causes them, the common types you will see, what a healthy exception rate looks like, and how to handle and reduce them. It is written for AP managers, controllers, and finance teams who want fewer held invoices and a faster close, not a software pitch.

What is an invoice exception?

An invoice exception is an invoice that cannot be processed automatically because its details fail a validation check or do not match the related purchase order, goods receipt, or contract. Instead of flowing straight to approval and payment, it is held and routed to a person to review and fix. Common triggers are a price that does not match the PO, a quantity that does not match what was received, or a missing PO number.

Think of it as the AP equivalent of a flagged transaction. The system is not rejecting the invoice; it is saying it cannot be sure the invoice is correct on its own. Until someone resolves the discrepancy, the invoice sits in a queue. Every invoice in that queue is a payment that is not happening yet and a vendor relationship that is slowly getting tested.

What causes invoice exceptions?

Invoice exceptions are caused by data that does not agree across documents or by information that is missing entirely. The vast majority are created before the invoice ever reaches AP, either by the vendor billing incorrectly, by a buyer who never raised a PO, or by data captured wrong at intake. They are rarely caused by the processing step itself.

Missing or invalid PO number

If an invoice has no purchase order reference, or the PO number is wrong or closed, the matching engine has nothing to compare against. The invoice drops to a queue until someone finds the right PO or confirms it is a non-PO invoice that needs manual coding and approval.

Price mismatch

The unit price on the invoice does not equal the price on the purchase order. This can be a vendor billing an old price, a negotiated discount the supplier forgot, or a genuine billing error. Price mismatches need a commercial review, not just a data fix.

Quantity mismatch

The quantity billed does not match the quantity recorded on the goods receipt. You ordered 100, the warehouse received 90, but the invoice bills for 100. The system holds the invoice rather than pay for items that were never received.

Missing or wrong data from capture

When the invoice number, date, vendor, or line amounts are read incorrectly at intake, the invoice fails validation even though the paper document is fine. Manual keying has an error rate high enough that this alone drives a meaningful share of exceptions, which is why clean extraction matters so much. When a field comes through blank or garbled, our guide to fixing missing invoice data covers how to recover it without re-keying the whole document.

Duplicates, tax, and coding errors

Duplicate submissions, tax amounts that do not foot, and GL coding that does not fit the vendor or expense type all trip validation rules. Each one is a small problem on its own, but together they fill the exception queue. Duplicates in particular are worth stopping at intake rather than in the queue; our guide on how to prevent duplicate invoice payments covers the checks that catch them, and consistent GL coding removes the coding half of this bucket.

What are the most common types of invoice exceptions?

The most common invoice exception types are price difference, quantity difference, missing or invalid PO, item not on the purchase order, value exceeds an approval limit, and receipt mismatch. ERP and AP platforms label these as payment blocks or exception codes, and each type routes to a different owner for resolution.

Price and quantity exceptions usually go to procurement or the buyer who owns the PO, because they need a commercial decision. Missing PO and item-not-on-PO exceptions go back to the requester to confirm the purchase. Value-exceeds-limit blocks go to a higher approver. Sorting exceptions by type, rather than treating them as one undifferentiated pile, is the single biggest lever for clearing them faster.

What is a good invoice exception rate?

A good invoice exception rate is around 10 to 15 percent, which is what teams with automated matching and clean capture tend to hit. Manual AP environments commonly run 25 to 30 percent. The exact target depends on how many of your invoices are PO-backed, but anything above 20 percent usually points to upstream data or process problems worth fixing.

To put it in volume terms, a team processing 500 invoices a month at a 20 percent exception rate is hand-touching 100 invoices every month. At three quarters of an hour each, that is roughly two full working weeks of staff time spent on rework instead of useful analysis. Cutting the rate from 20 percent to 12 percent gives most of that time back.

How do you handle invoice exceptions?

You handle invoice exceptions by identifying the exception type, routing it to the right owner with the supporting documents attached, resolving the discrepancy, and recording the root cause so it does not repeat. A consistent workflow with deadlines per exception type clears the queue far faster than ad hoc email chains.

1. Classify the exception

Tag each held invoice by type the moment it is flagged: price, quantity, missing PO, coding, duplicate, or approval. Classification decides who fixes it and how urgent it is.

2. Route it to the owner

Send price and quantity exceptions to the buyer or procurement, missing-PO exceptions to the requester, and over-limit exceptions to the right approver. Attach the invoice, PO, and receipt so the owner can decide without hunting for documents.

3. Set a deadline by type

Give each exception type an explicit service level, for example 24 hours to supply a missing PO and 48 hours to resolve a price mismatch. Exceptions without a deadline drift, and drifting exceptions become late payments.

4. Resolve and record the cause

Once the discrepancy is settled, release the invoice for payment and log why it happened. If one vendor causes the same price mismatch every month, the fix is a conversation with that vendor, not another month of manual rework.

How do you resolve a price or quantity mismatch?

You resolve a price or quantity mismatch by comparing the invoice against the purchase order and goods receipt, confirming which document is correct, and either correcting the invoice, updating the PO, or requesting a credit from the vendor. The cheapest time to catch these is at matching, before any payment is scheduled, when the correction cost is close to zero.

For a price mismatch, check whether the PO price reflects the latest agreed terms. If the vendor billed an outdated price, request a corrected invoice or credit. If the PO is stale, update it. For a quantity mismatch, trust the goods receipt: pay for what was received, and either short-pay with a clear note or hold the line until the rest of the shipment and its receipt arrive.

What is invoice exception management?

Invoice exception management is the systematic approach to identifying, resolving, and reducing the invoices that cannot be processed automatically. It covers the rules that flag exceptions, the workflow that routes and clears them, the metrics that track them, and the root-cause work that prevents them. The goal is fewer exceptions over time, not just faster firefighting.

Teams that treat exceptions as a managed process, with an owner, deadlines, and a monthly review of the top causes, steadily shrink their exception rate. Teams that treat each exception as a one-off surprise stay stuck at the same rate year after year. Fixing exceptions downstream without addressing the upstream data problem is a loop that never ends.

Why do invoice exceptions cost so much?

Invoice exceptions cost so much because each one takes roughly 45 to 60 minutes of skilled AP time to investigate, chase, and resolve, and because they cluster on a small share of invoices that absorb most of the team's day. Studies repeatedly find AP staff spending well over half their time on exceptions rather than value-added work.

The visible cost is labor, but the hidden costs are bigger: late-payment fees, lost early-payment discounts, strained vendor relationships, and a close that drags because invoices are stuck in review. A single missed 2 percent discount on a large invoice can dwarf the hourly cost of fixing the exception that caused the delay.

How do you reduce invoice exceptions?

You reduce invoice exceptions by attacking their causes upstream: enforce PO numbers on purchases, share buying terms with vendors so they bill correctly, capture invoice data accurately at intake, and set matching tolerances so trivial differences clear automatically. Prevention beats resolution because most exceptions are created before the invoice arrives.

Practical moves that work: require a valid PO before a purchase is approved, send vendors a short billing guide with your PO format and remit-to rules, set a small tolerance (say 1 percent or a few dollars) so rounding does not trigger a hold, and clean up duplicate vendor records that scatter history. Track your exception rate monthly and fix the top one or two causes each quarter rather than chasing every exception equally.

How does accurate data extraction reduce invoice exceptions?

Accurate data extraction reduces invoice exceptions by removing the data-quality errors that get introduced when fields are keyed by hand. When the invoice number, vendor, dates, line items, quantities, and amounts are captured correctly the first time, the matching engine compares clean data against the PO and receipt, so fewer invoices get flagged for differences that were never really there.

This is the part AP teams control before matching even starts. Tools like our invoice data extraction software read header and line-item fields from PDFs and scans and output structured data, so the figures going into your three-way match are the ones actually printed on the invoice. Because it reads by meaning rather than a fixed layout, it avoids the misreads that rigid template-based invoice parsing software produces on unfamiliar vendors. Clean capture will not fix a vendor who bills the wrong price, but it does eliminate the large bucket of exceptions caused by misread or mistyped data. For the ones that are a genuine pricing or quantity disagreement, the evidence trail you need is set out in our guide to resolving vendor payment disputes.

From there, the discrepancies that remain are genuine commercial issues that belong in your invoice processing software or accounts payable automation software, where matching tolerances and approval routing decide what clears and what needs a human. If you want the resolution side automated, tools that route held invoices through approval and payment close the loop, and pulling invoices straight out of your inbox with an email-to-data parser keeps PO references from getting lost in forwarded attachments. To see how matching itself works, our guide to three-way matching covers the check that most exceptions are flagged against.

Exceptions will never hit zero, and they should not: the held queue is doing its job when it stops a wrong payment. The win is making the queue small and fast, so the handful of invoices that genuinely need a human get one quickly, and everything else just pays on time.