Purchase Orders

Why purchase orders are critical

A project’s true financial position is three lines, and only three: budget, then committed — the value of approved purchase orders and subcontracts — then actual, meaning invoiced or paid. Forecast final cost is actuals, plus open commitments, plus an estimate of the balance still uncommitted.

Firms that track only actuals discover their overruns 30 to 60 days late. That is not a failure of diligence; it is the invoice lag, and it is structural. The liability already exists by the time the invoice reports it, and every option that might have been taken has expired.

The leak has a name: maverick spend. Cross-industry research puts roughly 30% of indirect spend off-contract — as much as 80% at the worst performers — forfeiting 10 to 20% of negotiated savings in the process. Construction’s version is entirely familiar: the site engineer phones a supplier for urgent materials, and the invoice arrives with no PO to match it, no agreed price, no committed cost, and a dispute already inside it.

Only about a quarter to a third of construction projects finish within 10% of their original budget. Commitment-stage control is the most under-used lever on that number.

The role of POs in project performance

• Put the scope in the right instrument. A purchase order suits defined-scope supply and simple services, on the PO’s own terms. A scope that needs supervision, insurances, retention or defects-liability obligations belongs in a subcontract. Labour scopes misplaced onto material POs is a real audit finding, and the reason is not pedantic: the retention and the insurance obligations silently vanish with the paperwork. Framework agreements sit above both — pre-negotiated rates drawn down by call-off orders — and they carry value caps and expiry dates that have to be tracked, or the ceiling gets quietly exceeded by people who had no idea there was one.

• Revisions are commitments too. Every change to quantity, price or date is a numbered revision with an approval behind it, and the cost report must always reflect the latest approved revision. A cost report showing Rev 0 while Rev 3 is in force is not being conservative. It is fiction with a delay.

• Partial deliveries make a PO a cumulative document. One order for 500 tonnes of rebar is fulfilled across weeks of deliveries. Goods receipts draw down the PO line as they arrive; over-delivery tolerances — say ±5% on bulk materials — are agreed in advance rather than argued afterwards; and the invoice matches against cumulative receipts, never against the raw PO quantity. Match against the PO total and you will pay for steel that has not turned up.

• Closeout is monthly hygiene. Purchase orders left open with undelivered residual balances hold phantom commitment against the budget. Closing them releases real money. And the inverse failure is just as expensive and much less obvious: believing the budget is exhausted when it is not, and running a value-engineering exercise against a number that was never true.

International orders carry structural terms

Gulf contractors import heavily, and on an import the PO is carrying more than a price.

The Incoterm decides who carries freight, marine insurance and customs clearance — and, critically, when risk transfers. FOB, CIF and DDP are not stylistic preferences; they are different allocations of cost and of risk, and getting one on the PO by habit rather than by decision is how a delivery becomes somebody’s expensive surprise.

The currency matters too. A euro-denominated purchase order sitting against a dirham budget means your committed cost moves with the exchange rate, whether or not anyone is watching it. And customs and legalisation need real lead-time buffers rather than optimistic ones. On long-lead equipment, advance payments ride on advance payment guarantees — which is where the purchase order stops being a procurement document and starts being a contract-management one.

What happens without PO discipline

The chains are short. A verbal order, then an invoice at a price nobody agreed, then a disputed and unaccrued liability, and a margin surprise at close. A revision never captured, so the project reads as on budget right up until the Rev 3 invoice lands and the forecast’s credibility goes with it. Stale POs never closed, so committed cost is overstated and the team manufactures a budget crisis that does not exist.

Every one of those is a workflow failure wearing the costume of a cost overrun.

And the realistic fix for verbal orders is not policing. It is speed. If raising a proper PO takes longer than a phone call plus a favour, the phone call wins — every time, on every site, no matter what the procedure says. An emergency PO issued in ten minutes beats a verbal order comfortably. The process has to make the disciplined path the fast path, because the disciplined path will not otherwise be taken.

How Zepth runs purchase orders

Every PO — material or service — carries its line items, terms, currency and delivery data, flows through amount-threshold approvals set by your delegation of authority, and posts to committed cost the moment it is approved. Not at invoice. At approval, which is when the money actually became owed.

Revisions are numbered, approved and audit-trailed. Deliveries draw down the lines. Invoices match against receipts. And the budget–committed–actual picture is live at both project and portfolio level, so the forecast is a screen rather than a month-end excavation. An award from a tender converts directly into the PO, so nothing is re-keyed between what was bid and what was bought.