The Real Cost of RTO and Failed Deliveries

A return-to-origin parcel is the most expensive event in fulfilment, and the one most likely to be recorded as a general shipping cost rather than attributed to anything.

It is worth separating from a customer return first, because they are different events with different fixes. A customer return means the goods arrived and were rejected — a product, sizing or expectation problem. An RTO means the goods never arrived at all — an address, availability, payment or intent problem. Merging them in reporting hides both.

What it actually costs

Freight in both directions is the obvious part. The rest adds up to as much again:

• Outbound label, already paid the moment the parcel was collected.
• Return leg, billed separately, often under a different service code and frequently at a similar rate.
• Packaging, consumed. Outer carton, void fill, tape, insert.
• Handling and restocking — someone opens it, inspects it, puts it away.
• Failed-attempt or COD handling fees, depending on the carrier contract.
• Markdown, where the item comes back damaged or out of season.

Cost of goods is the one thing recovered, provided the item returns sellable. That is why the loss is $23.05 rather than $92 — but $23.05 against a $46.25 contribution margin means one RTO undoes half a successful order.

The rate is the number, not the incident

Individual RTOs feel like bad luck. The rate is a structural property of how you sell.

The arithmetic is straightforward: at a ten percent RTO rate, every ten attempted orders produce nine successes and one failure. Nine orders contribute $416.25 between them and the failure takes back $23.05 — about five and a half percent of the margin, gone before any other cost is considered.

At twenty percent it is over eleven percent of contribution margin. For most stores that is larger than their entire net margin.

This is why RTO belongs in the acquisition calculation rather than in a fulfilment report. You pay to acquire attempted orders, and expected margin per attempt is what a customer is actually worth.

Where it concentrates

RTO clusters, and finding the clusters is most of the fix.

Postcode. The dominant cut in most markets. Address quality, carrier coverage and local delivery practice vary enormously, and a small number of areas usually account for a disproportionate share.

Payment method. Cash on delivery fails far more than prepaid, because refusing at the door costs the customer nothing.

Order value. High-value COD orders fail more — more cash to produce, more room for second thoughts.

Acquisition channel. Casual traffic from aggressive discounting converts casually and refuses casually.

Time to dispatch. Parcels that take longer to arrive fail more often, because enthusiasm decays and circumstances change.

Cut by postcode and payment method first. Between them they usually explain most of the variation, and both have direct interventions.

Attribution, and why it matters here

RTO cost has to be attributed back to the originating order, the product and the region, not booked as a general fulfilment expense.

Booked generally, it is a cost line that goes up and down for no visible reason. Attributed, it becomes actionable: this product line has a fifteen percent RTO rate, these postcodes account for a third of failures, this channel’s customers refuse three times as often as that one’s.

None of those findings survive averaging. And because RTO freight arrives on a carrier invoice weeks later, keyed by tracking number rather than order number, the attribution has to be deliberate — the join is available and nobody makes it by accident.

Putting it into the acquisition ceiling

The reason to cost RTO precisely is that it changes what you can afford to pay for a customer, and most stores never make that connection.

Acquisition cost is paid on attempted orders. Contribution margin is earned on delivered ones. So the figure to compare against CAC is expected margin per attempt, which folds the failure rate in:

(delivery rate × margin when delivered) − (failure rate × cost of a failure)

At a ten percent RTO rate on the numbers above, a $46.25 contribution margin becomes $39.32 in expectation. A store bidding against the higher figure is overpaying by nearly seven dollars on every customer, and the error is invisible because the ad account and the fulfilment report never meet.

The same adjustment should flow into any channel or region comparison. A region with a fifteen percent RTO rate needs a materially lower acquisition cost than one at three percent to produce the same profit — and judged on delivered-order margin alone, the two regions look identical.

The interventions, ranked by evidence

Partial prepayment on COD orders. The strongest single lever in markets with high COD share. It converts a costless refusal into one the customer has already paid to avoid.

Order confirmation before dispatch. Effective, cheap, and it also catches address errors before freight is spent.

Address verification at checkout. Removes the subset of failures caused by bad data rather than intent, which in some markets is most of them.

Postcode restrictions or carrier switching for persistently poor areas. Precise, and the regional cut tells you exactly which list to build.

Faster dispatch. Indirect, and it works — parcels that arrive sooner fail less.

Each trades some conversion for a lower failure rate, and each trade is measurable against the $23.05 figure. That is the point of calculating the cost of one RTO properly: it turns an argument about customer experience into an arithmetic comparison.