Invoice the day you deliver: the shortest lever on cash flow

Payment terms are negotiated. Someone in your office pushed for 30 days and settled at 45, and there is a contract to prove the argument happened. The gap between the day you deliver and the day you invoice is negotiated by nobody. It is given away a few days at a time, and it costs exactly what those 45 days cost.

The mechanism is worth stating precisely, because it is usually described loosely. Under the EU Late Payment Directive (2011/7/EU) — and the Swedish räntelagen that implements it — statutory late-payment interest between businesses runs from 30 days after the invoice reaches the customer, where no other term is agreed. It does not run from the day the goods arrived. Deliver on the 3rd, invoice on the 17th, and those fourteen days are not late, not overdue and not accruing anything. They are a loan you made, interest-free, without being asked and without being thanked.

That is why this is the shortest lever available. Shortening payment terms means renegotiating with a customer who has no reason to agree. Shortening the delivery-to-invoice gap requires nobody’s permission.

Where the days actually hide

Almost nobody delays an invoice on purpose. The delay is assembled out of three waits, and they have different fixes.

Waiting for the proof of delivery. If the signature is on paper, it comes back at the speed of the vehicle. Not one day — the remainder of the tour, plus the trip through the office, plus the time it spends in a pile before someone matches it to a booking. A signature captured on a phone at the dock is in the system before the driver has left the yard. We covered this at length in the article on paper PODs; it is the largest of the three for most carriers, and the most visible.

Waiting for the last uncertain line. Waiting time at the ramp, a toll that has not appeared yet, an extra stop the driver called in from the road. The invoice is complete except for one figure, so the whole invoice waits. This is the quietest of the three, because from the inside it looks like diligence.

Waiting for the batch. Invoicing on Fridays. Invoicing at month-end. Nothing is wrong and nothing is missing — the work is queued because queuing it is the habit.

The batch is the biggest and the least examined

This one is arithmetic rather than opinion. If your deliveries are spread across the month and every invoice goes out on the last day, the average delivery-to-invoice gap is around fifteen days — before anyone has made a mistake, lost a document or fielded a query. A perfectly run month-end batch still builds in half a month. No amount of diligence inside the batch removes it, because the delay is the batch.

For cross-border work there is a compliance edge to this as well. Swedish VAT rules set an actual deadline for some transactions: an intra-EU supply of goods to a VAT-registered buyer in another member state, and main-rule services to EU buyers, must be invoiced by the 15th of the month following supply. Ordinary domestic sales have no statutory deadline — Skatteverket points at god affärssed, good business practice — but a monthly batch that slips turns a habit into a reporting problem for the cross-border half of the work.

It is worth noting what is not coming to help. The Commission’s 2023 proposal to replace the Directive with a Regulation capping all commercial terms at 30 days is, as of mid-2026, blocked in the Council. Plan on the rules you have.

What invoicing on the delivery day actually requires

Five things, and none of them is a setting you switch on.

1. The POD has to arrive electronically. Signature, photos and any failed-delivery reason, captured at the stop and attached to the booking with no human step in between.
2. The price has to be resolved before the job, not after it. If working out what to charge means finding an email and remembering an agreement, invoicing will always be a research task. Agreed rates held as price sheets — by weight, distance, time or zone — make the charge something the system computes rather than something a person reconstructs.
3. You need a rule for the uncertain line. Decide in advance whether an unresolved toll blocks the invoice or is billed separately later. Either answer is defensible. Having no answer means every exception is debated individually, at the cost of everything queued behind it.
4. Someone has to own the exception queue. Same-day invoicing turns a weekly pile into a daily trickle. That is less work in total, but it is work that needs a name against it, or it quietly becomes a weekly pile again.
5. The accounting handoff has to be a sync, not a retype. Rekeying is both a delay and an error source, and the errors cost more than the delay — see accounting integration basics.

What it costs, honestly

Faster invoicing is not free, and three trade-offs are worth facing before you commit to it.

Invoicing fast means invoicing wrong faster, if the data is not there. A disputed invoice is paid later than a correct one sent a week after delivery, and it spends an argument with a customer as well. Accuracy is a precondition here, not a competing objective. If your charges are not reliably right at the moment of delivery, fix that first and the speed follows.

More invoices is not automatically the goal. Some customers genuinely want one consolidated invoice a month, and forcing per-delivery invoices on them can slow payment rather than speed it, because it fights their own approval process. The lever is the lag, not the frequency. If a customer wants one invoice a month, the win is issuing it on the 1st instead of the 12th.

A correction has a real cost. An issued invoice is an accounting document; in Sweden you do not edit it, you credit it and reissue. That is the correct process, but it is a second document and a second conversation — so the threshold for issuing early should be genuine confidence, not optimism.

Measure two numbers, not one

Most carriers watch days-to-cash as a single figure, which is exactly why this lag stays invisible: it is buried inside a number that is mostly about payment terms and chasing. Split it.

Measure delivery to invoice on its own. Measure invoice to payment separately. They have different owners, different causes and different fixes, and only the first is entirely within your control. Then look at the tail rather than the average — the median is usually respectable, and the money is in the 10% of jobs that took three weeks because a document went missing. That tail is a list of specific bookings someone can go and read, which is a far more useful artefact than a percentage.

Where navichain fits

navichain closes the delivery-to-invoice gap where it starts: signed proof of delivery is captured on the driver’s phone and sealed, hashed and timestamped at the stop, so the office is not waiting for paper to travel. Charges are computed from your own price sheets by weight, distance, time or zone; you can invoice a single booking, merge several, or bill a whole period in one pass; issued invoices are sealed, and cancelling one raises a credit note rather than rewriting history; and finished invoices sync to Visma, QuickBooks or Tripletex instead of being retyped. Customers pull their own documents, PODs and invoices from the portal. The platform page shows what is inside, and pricing is public, from 995 kr a month with every feature on every plan.