Three Way Matching for Offshore AP Exception Teams in India

How many of your supplier invoices clear three way matching in accounts payable on the first pass, and how many land in a queue for a person to sort out?

This guide is for finance and operations leaders at US and UK companies who have already bought the automation and are still looking at an exception queue every morning.

We help global companies hire accounts payable exception analysts in India through our Employer of Record service, so this guide focuses on the judgment work that matching hands back to people rather than on which tool to buy.

You will get the mechanics of the match, the exceptions that actually recur, how to set tolerances that hold up, and how to staff the queue. It sits alongside our wider guide to the automated invoice processing workflow, which walks the full invoice lifecycle rather than the match itself.

What is three way matching in accounts payable?

Three way matching in accounts payable is a control that compares three documents before an invoice is approved for payment: the purchase order, the goods receipt, and the supplier invoice. If item, quantity, and price agree across all three within tolerance, the invoice posts. If any one disagrees, it stops.

Each of those documents is created by a different party at a different moment, and that separation is the whole point.

Purchasing raises the order. Receiving confirms what arrived. The supplier tells you what to pay. Agreement between three independent records is far harder to fake or fumble than agreement between two.

It is worth being precise about what each document proves, because most match failures come from asking a document to prove something it never could:

Read that way, the match is not an accounting chore. It is the last place a wrong price, a short delivery, or an invoice for something you never ordered can be caught before the money leaves.

It is the control step inside the wider procure to pay process, and it generates almost all of the manual work that follows.

Most companies bolt matching onto an existing setup rather than design it, which is why accounts payable automation projects so often stall at exactly this stage.

So what does the match actually do, in order?

How does a three way match run step by step?

The system pulls the purchase order referenced on the invoice, finds the goods receipts posted against it, and compares quantity, unit price, and line item across all three records. Anything inside your tolerance rules posts automatically. Anything outside them is held and routed to a person.

The sequence matters more than most teams expect.

In practice the match runs through a fixed order of checks, and an invoice can fail at any one of them:

• Purchase order lookup: The invoice must carry a valid PO number that exists and is still open. A missing or closed PO stops the match before it starts.
• Line mapping: Invoice lines are mapped to purchase order lines. Suppliers who combine, split, or reorder lines break this quietly.
• Quantity check: Invoiced quantity is compared to the quantity received, not the quantity ordered. Partial deliveries are the usual source of trouble here.
• Price check: Unit price on the invoice is compared to the purchase order price, and in stronger setups to the contract rate card behind it.
• Extended value check: Quantity multiplied by price is tested against the invoice total, which catches rounding, currency, and tax errors.
• Tolerance test: Each variance is measured against the tolerance rules you set, by percentage, by absolute value, or by both together.

An invoice that passes all six posts without a human touching it. An invoice that fails any one of them becomes somebody's morning.

Companies running source to pay operations in India usually find the first two checks cause more failures than the price and quantity checks everyone worries about.

What happens to invoices with no purchase order?

Services, utilities, professional fees, and rent often arrive without a purchase order, so there is nothing to match against.

These route down a different path, usually a coding and approval workflow rather than a match. They need a named owner or they sit unapproved for weeks and surface as a surprise at close.

What happens to goods received with no invoice?

The opposite gap is goods received but not invoiced, which sits on your balance sheet as an accrual and grows quietly while nobody owns it.

Somebody has to chase the supplier, confirm the receipt was genuine, and clear or reverse it. That work belongs with the same team that owns supplier risk management, because a supplier who never invoices is often a supplier with a problem.

Which brings us to the exceptions themselves.

What are the most common three way match exceptions?

The recurring ones are quantity variances from partial or over deliveries, price variances from stale purchase orders, missing or wrong PO references, receipts posted late or not at all, and unit of measure mismatches. Almost all of them are data problems created upstream, not payables mistakes.

From our experience helping companies build finance teams in India, the exception queue reads as a report card on procurement and receiving rather than on payables.

It is worth sorting them by where the fault actually sits, because that tells you who has to fix it and whether it will come back next month:

Notice how few of those are payables errors. The exception team spends most of its day resolving other functions' data, which is exactly why the role needs standing and authority as well as accuracy.

The same logic drives account reconciliation software selection. The tool clears what agrees, and the value sits entirely in how quickly a person resolves what does not.

At period end the unresolved queue becomes a close problem, and financial consolidation software simply inherits whatever payables did not finish.

Most of those exceptions are governed by one setting that almost nobody revisits.

How do you set match tolerances that catch real errors?

Set tolerances by risk rather than by convenience. Use a percentage band for high value lines and an absolute currency floor for low value ones, so a small percentage on a large invoice cannot pass unseen and a trivial rounding difference on a small one does not consume an analyst's hour.

A single flat tolerance applied to every supplier and category is the most common design error we see.

Set it tight and you drown the team in noise. Set it loose and you approve real overcharges, invisibly, for as long as the rule stands.

A workable tolerance design separates four decisions that are usually collapsed into one:

• Variance type: Quantity, price, and extended value each deserve their own tolerance. They fail for different reasons and carry different risk.
• Direction: Under invoicing rarely needs the same scrutiny as over invoicing. Many teams let favourable variances through automatically and log them.
• Basis: A percentage band and an absolute floor used together, with the invoice held only when it breaches both, keeps small noise out of the queue.
• Scope: Tolerances set by supplier tier or spend category reflect real risk far better than one global rule ever will.

Write the rule down, date it, and review it after a quarter of live exception data. Tolerances set on day one from guesswork are almost always wrong, and almost never revisited.

Treat this as a control rather than a configuration setting. It belongs with the rest of your finance automation controls and should be documented and reviewed the same way.

Companies that already run accounting outsourcing to India usually hand tolerance monitoring to the offshore lead, because that is the person who sees the pattern shift first.

None of that answers the question buyers actually keep asking.

What can three way matching software not do on its own?

Software compares documents and applies rules. It cannot decide whether a price increase was agreed, whether a short delivery is acceptable, whether to pay a disputed invoice to keep a supplier shipping, or whether a run of small variances is sloppiness or fraud. Those are all judgment calls.

This is the part of the product brochure that does not exist.

Here is the honest split between what a rules engine handles and what a person still has to decide:

Every row in that right hand column is a person. As the software improves, the left column grows and the right column gets harder, because only the genuinely ambiguous work survives.

That is the honest shape of AI in accounts payable today, and it is why headcount plans that assume a straight line reduction tend to disappoint.

We have written more broadly about what stays human on an AI-augmented offshore team, and payables is one of the clearest examples of it.

If the remaining work is judgment, the measurement has to change too.

How do you measure whether your matching process is working?

Track first pass match rate, exception ageing, exceptions grouped by root cause, and rework. Judge each against your own baseline month to month. What matters is whether last quarter's most common exception is smaller this quarter, and whether anything is ageing past your payment terms.

Resist the urge to benchmark against a published figure you cannot verify. Your mix of PO and non-PO spend alone makes those comparisons close to meaningless.

Four measures tell you almost everything about how a matching operation is running:

• First pass match rate: The share of invoices clearing every check with no human touch. Movement in this number is the single best signal you have.
• Exception ageing: How long held items sit unresolved. A stable queue size can still hide a growing tail of old and difficult items.
• Root cause mix: Exceptions grouped by cause rather than by count. This is the view you take to procurement and receiving.
• Rework rate: How often a resolved exception comes back. A high rework rate means the fix was cosmetic and the cause is untouched.

Report the root cause mix upward every month. It is the only one of the four that changes another team's behaviour.

Receivables works the same way, where days sales outstanding is a single number that only means something measured against your own history.

Teams that run payables and the order to cash process from the same offshore hub usually standardise the reporting cadence across both, so one monthly pack covers the whole working capital picture.

The tooling question mirrors it too, which we cover in our guide to accounts receivable software.

Measurement only helps if somebody owns the number, which is a staffing question.

Which roles does an offshore AP exception team need?

You need five: an exception analyst who works the queue, a supplier master data owner who fixes what causes it, a procurement liaison who chases purchase order amendments, a controls and compliance reviewer, and a team lead who owns tolerances, escalation, and the monthly root cause report.

These are genuinely separable jobs. Collapsing them into one capable generalist is why so many offshore payables teams plateau after the first six months.

Here is what each one owns day to day:

• AP exception analyst: Works the held queue, resolves quantity and price variances, and releases what falls inside delegated authority.
• Supplier master data owner: Fixes the item, unit of measure, tax, and banking records that cause repeat exceptions, and controls who is allowed to change them.
• Procurement liaison: Chases purchase order amendments, confirms whether a price increase was agreed, and closes the loop back to the buyer who raised it.
• Controls and compliance reviewer: Checks segregation of duties, samples released exceptions, and prepares the evidence pack auditors ask for at year end.
• AP team lead: Owns tolerance settings, escalation thresholds, the monthly root cause report, and the working relationship with the onshore controller.

Staffed that way, the team absorbs the exception queue and then starts shrinking it at source, which is the outcome no tool delivers on its own.

The seniority mix and the full cost picture for these roles are set out separately in our guide to building an offshore accounts payable team in India, so this article deliberately does not repeat the numbers.

The wider function these roles sit inside is covered in our guide to offshore finance and accounting in India.

For budgeting across the whole department rather than one process, see the cost of an offshore finance team in India.

Companies that already run an offshore FP&A team often add payables next, because the reporting rhythm and the review cadence are already in place.

Wisemonk supports 300+ global clients, more than 2,000 employees, and over $20M in annual payroll, with a 4.8/5 rating on G2.
Wisemonk, 2026

At that scale our answer to staffing questions is usually about sequence rather than size. Start with the lead and one analyst, then add where the root cause report says the volume actually is.

The mechanics of building an offshore team in India are the same for payables as for any other function.

Which leaves the buying decision on the software side.

What should you ask before you buy a matching tool?

Ask how the tool handles your worst invoices, not your best. Ask what happens to non-PO spend, how tolerance rules get configured and changed, who can override a held invoice, and what the total cost includes once implementation, integration, and per document charges are counted.

Matching and AP automation are quote-based categories. Any list price you find is a starting point for a negotiation you have not had yet.

Rather than chase a headline number, price the components and take the questions into the conversation:

Those seven answers will tell you more about total cost than any headline figure, and they surface the work that quietly lands back on your own team.

The build versus buy question behind it looks a lot like the EOR vs entity in India decision, where the more affordable option on paper is often the one that hands you the most operational work.

If you are still weighing where this work should sit at all, our guide to outsourcing to India covers the engagement models and what each one really costs to run.

And our guide to offshoring to India covers what changes when the team is yours rather than a vendor's, which matters a great deal for a control like this one.

Through an Employer of Record, a compliant offer can be issued in 24 to 48 hours, and hiring an Indian national typically takes 1 to 2 weeks.
Wisemonk, 2026

Location matters less for payables than most buyers expect, though it is worth reading our view on the best Indian cities for offshore finance operations before committing to a hub.

Day to day, the difference between a queue that shrinks and one that grows is management attention, which we cover in our playbook on managing offshore teams in India.

Here is where we fit into that.

How can Wisemonk help you build three way matching operations in India?

Wisemonk is an India-native Employer of Record (EOR) that helps global companies hire, pay, and manage talent in India without setting up a local entity.

For three way matching, that means an exception analyst or a full payables pod at their desk within weeks, on compliant Indian employment contracts, without registering a company in India first.

We support 300+ global clients, more than 2,000 employees, and over $20M in annual payroll, and we are rated 4.8/5 on G2. EOR pricing starts at $99 per employee per month, as of August 2026.

Here is how we help:

• Recruitment: We source and screen exception analysts, supplier data owners, and team leads who have worked a matching queue before.
• Managed payroll: We run monthly salary, statutory deductions, and filings for your India payables team.
• Contractor management: We contract and pay specialists compliantly when you need short-term cover through a close or a system migration.
• Background checks: We verify employment and education history before anyone gets access to your supplier master and payment approvals.
• GCC setup: We help you grow a payables pod into a wider capability centre once the headcount justifies it.
• Entity setup: We handle Indian company registration when you decide to move from an EOR to your own entity.

From our experience staffing payables exception queues in India, the hire that moves the numbers fastest is not another analyst but the supplier master data owner, because one corrected unit of measure removes an exception that would otherwise recur on every future invoice from that supplier.