What is a purchase order (PO)? A complete guide for project-based businesses

A purchase order is a document a buyer sends to a supplier to confirm an order of goods or services at an agreed price and terms. Once the supplier accepts it, that document becomes a binding agreement — and, for a business running projects, the moment a cost becomes real.

If you're ordering materials for a build, hiring a subcontractor, or chartering equipment for a refit, you've raised a purchase order — whether you called it that or not. This guide covers what a purchase order is, what goes on one, how the purchase order process works end to end, and why the timing of a PO matters more than most teams realise when they're trying to keep a project on budget.

What is a purchase order?

A purchase order, usually shortened to PO, is a formal document a buyer issues to a supplier to place an order. It records what's being bought, how much, at what price, and when it's due for delivery. In legal terms it's an offer to buy on stated terms; once the supplier accepts it, the PO becomes a binding agreement between the two parties.

That's the definition. The useful part, though, is what a PO does that a casual email order does not: it creates a written, numbered, referenceable record of a commitment that both sides — and your own finance team — can point to later.

A purchase order is an offer to buy on stated terms. It becomes binding when the supplier accepts it — which is why the approved PO, not the later invoice, is the first honest signal that money is committed.

What goes on a purchase order

Most purchase orders carry the same core fields, regardless of industry or business size:

On a project, the total value field is the one to watch. A single subcontractor PO on a construction job can run to six figures — and the day it's approved is the day that money is spoken for, even though no invoice will land for weeks. If you're setting up a PO for the first time, most teams start with a simple format in Word or Excel

A purchase order example

Here's what one looks like in practice — an engineering consultancy issuing a PO to a specialist subcontractor for a piece of structural analysis on an industrial build:

Every field on that document does a job. The PO-2847 reference is what the supplier will quote on their invoice so you can match it later. The line items and totals are the commitment against the project budget the moment approval lands. The payment terms set when the invoice is due, once delivered.

The purchase order process, step by step

The purchase order process is the path a single order takes from first request to final payment. It's worth seeing the whole sequence, because the point where most teams lose visibility isn't where they expect.

1. Purchase request raised. Someone on the team asks internally to buy something — this is a request, not yet an order.
2. Approved & PO number assigned. Whoever owns the budget approves the request. A PO number is assigned, the document becomes a formal purchase order, and the cost is committed against the project.
3. PO sent to supplier. The approved PO goes to the supplier, who accepts it and starts work.
4. Goods or services delivered. The supplier delivers what was ordered, ideally with a delivery note or goods-received record.
5. Invoice received. The supplier bills for the delivered work.
6. Three-way match. Before paying, you compare the PO, the delivery record, and the invoice, and check they agree.
7. Payment. Once the three line up, the invoice is approved and paid.

Here's the timing problem. In a lot of project businesses, finance only sees a cost at step 5 — when the invoice arrives. But the money was committed at step 2, sometimes weeks earlier. On a single small order that gap doesn't matter. Across dozens of open POs on an active project, it's the difference between knowing where your budget stands and finding out after the fact.

How to raise a purchase order

The mechanics of raising a purchase order are straightforward, and they're the same whether you're using a spreadsheet or a dedicated system. To raise a purchase order:

1. Confirm the need — what's being bought, for which project, and against which budget.
2. Gather the details — supplier, agreed quantities, unit prices, delivery date, payment terms.
3. Draft the PO, populate the fields (the ones listed above), and assign the next PO number.
4. Send for approval by whoever owns the budget for that spend.
5. Once approved, send the PO to the supplier and confirm they've received and accepted it.

What changes as you scale isn't the steps — it's who does them, and whether approvals happen fast enough to keep pace with the work.

Purchase requisition vs purchase order

These two terms get mixed up often, and it's worth being clear on the difference. A purchase requisition — sometimes called a purchase order request — is an internal request to buy something. It asks for permission. A purchase order is the external document that places the order once that request is approved. Requisition first, purchase order second.

Small teams often skip the requisition step entirely, because the person who needs the purchase is the person who can approve it. Larger teams, or teams with more people spending against shared project budgets, keep the requisition step in place so that spend gets checked before it gets committed. The full breakdown, and when to keep the requisition step in place, is in purchase requisition vs purchase order.

Purchase order numbers, briefly

Every purchase order gets a unique PO number when it's raised — for example PO-2847 or BLD-2026-0042. That number is what ties the whole life of the order together: the approval, the delivery note, and eventually the supplier's invoice all reference the same PO number, which is what lets you match them back to one commitment. We cover formats, generation, and where the number appears on an invoice in what is a PO number: how purchase order numbers work.

Why PO discipline matters for project budgets

For most businesses, a purchase order is admin — a document that precedes an invoice. For a project-based business, it's a control point. Here's why the difference matters.

On a live project, several people are usually spending against the same budget: a project manager approving subcontractor hours, a site lead ordering materials, a designer commissioning specialist input. Each purchase, individually, is reasonable. In aggregate, they either fit the budget or they don't — and by the time invoices arrive to make that visible, the decisions that caused the drift are already weeks old.

The PO is where that becomes fixable. If every purchase goes through a PO that's approved against the current committed total, the drift is prevented before it happens rather than reported after. That's what people mean by tracking committed cost across a project: not just tracking what's been invoiced, but what's been committed — the sum of approved POs that will become invoices later.

The approved PO is the earliest honest signal that budget has moved. Capturing cost at approval — not at invoice — is the whole idea behind real-time cost control.

Here's what that looks like on a live project:

The invoiced figure is what your accounting system already shows you. The committed slice — money that's spoken for but hasn't been billed yet — is the part that's normally invisible until too late. That's the piece a purchase order captures at the moment of approval, if you're set up to look.

Three-way matching, at the end of the process

Three-way matching is the check you run before paying a supplier. You put three documents side by side — the purchase order, the record of what was delivered, and the invoice — and confirm they agree on quantity and price. If the invoice bills for ten units and the delivery note shows eight, the PO is what tells you which is right.

A good cost system won't decide this for you, but it will show the three documents side by side so any discrepancy is easy to spot, rather than buried in a pile of paperwork on someone's desk. The full walkthrough is in what is 3-way matching.

Types of purchase order

Most orders fall into one of four types:

Purchase orders in construction

The reason PO timing matters so much comes down to the shape of the work. In construction, a project might carry dozens of open subcontractor and materials POs at once, each a sizeable commitment against a fixed budget, each raised by a different person. The same pattern shows up across the industries that live and die by project margin — engineering, marine, energy, and events: large commitments, made early, invoiced late.

Construction has its own vocabulary and its own supply-chain shapes on top of that — retentions, variations, subcontractor certification. For a deeper look at how POs work on a build, see purchase orders in construction.

When spreadsheets stop working — signs it's time for PO software

Spreadsheets and emailed approvals work until two things happen at once: the orders get bigger, and there are more of them open at the same time. That's the normal state of an active project. None of the following on its own is a crisis, but if several are true at once, it's usually the signal to move:

• Nobody's fully sure which POs are still open, or against which budget.
• A commitment gets approved that the budget couldn't really take, and it isn't caught until the invoice arrives.
• Finance is reconstructing what was spent from invoices, weeks after the decisions were made.
• Two people raise POs with the same number, or numbering skips.
• Approvals sit in inboxes until the work has already started.
• You can't answer "how much of this project's budget is already committed?" without opening several files.

None of that is a failure of effort — it's what manual tracking does when the volume climbs. Moving to a purchase order system is real setup work. What it buys you is the timing: commitments captured the moment a PO is approved, shown against the budget while there's still room to act. If you're weighing up options, purchase order software: what to look for walks through how to choose.