Tracing shipping costs back to margin: what a shipment really costs per order

Short answer. Tracing shipping costs back to margin means that for each order you assign the actual label cost, the correct carrier rate, any customs charges and packaging to that specific order, and you deduct that from the revenue per product and channel. Only then do you see your real margin, not an average that lies.

Shipping software buys a label. That is it. Sendcloud, MyParcel or the direct connection with PostNL and DHL print a sticker and book the cost. The real work sits in the decision that precedes it, and in what you do with that cost line afterwards. I sold cross-border for years through my own webshop, bol and Amazon, and the most expensive mistake I made was treating shipping costs as one pot. That average hides exactly the orders you lose money on.

Why an average shipping rate ruins your margin

A flat average of, say, 5 euros per order sounds convenient, but it is never right. A light parcel within the Netherlands costs something completely different from a fragile item to Germany with signature on delivery, or a two-kilo shipment to France. Assign that average to every product and your heavy and international orders look more profitable than they are, while your light orders look poorer. You are then steering on figures that do not exist.

The costs are also channel-dependent. On bol and Amazon the SLAs for delivery time are higher, so there you more often have to choose a faster and more expensive service. If you do not trace that back to the channel, you think bol is your best channel while the shipping and commission costs are eating the margin there the hardest.

The choice per order: carrier and service are a decision, not a setting

Which carrier and which service you choose depends on a few variables that differ per order: destination, weight, value, fragility, the channel's SLA and the price. Parcels over 30 kilos fall in a grey area where one carrier stops and another keeps going, and above 70 kilos you move to pallet freight (Sendcloud). A multi-carrier setup lets you choose, per shipment, the carrier best tuned to that destination and size (Picqer).

Sometimes splitting across two parcels is cheaper or safer than one heavy or fragile parcel. That is a calculation rule, not a gut feeling: compare the rate of one parcel in the high weight class with two in the lower class, including the higher risk of damage and returns. That trade-off should happen automatically the moment the order comes in, not by someone standing at the packing table guessing.

Examine the process first, then build

Before I build anything, I look at the process as it runs now. With resellers it has almost always grown organically. Someone picks carriers out of habit, splits parcels on instinct, and the shipping costs land somewhere in the bookkeeping without ever being fed back to the product. Part of those steps can be smarter, and part can simply go. Maybe you use three carriers where two would do, or there is a manual check that no longer catches any errors.

This is a deliberate step, not an excuse to build nothing. I strip out the ballast so I can then automate the real decision: choosing the right service per order and writing the actual costs straight back to the margin per product, channel and supplier.

What a shipment really costs: the cost line per order

The calculation that needs to happen is simple in form, hard in practice because the data comes from different systems. Per order you add up:

• the actual label rate of the chosen carrier and service
• packaging costs (box, filling, tape)
• any customs charges on shipments from outside the EU
• the part of the channel commission that touches shipping (on bol and Amazon)

You deduct that from the revenue, and you write it back to the product and channel line. If you want a rule of thumb for your free-shipping threshold: divide your shipping costs by your margin percentage and add that to your average order value, for example 5 euros divided by 0.40 is 12.50 euros extra (Webwinkelsucces). But a rule of thumb is no substitute for the real figures per order.

Customs rules change as of 1 July 2026

If you source or ship cross-border from outside the EU, something fundamental changes. The 150-euro de-minimis exemption for e-commerce shipments will be definitively abolished as of 1 July 2026, and in its place comes a fixed customs duty of 3 euros per product category per shipment (evofenedex). A parcel with two different HS codes can therefore cost 6 euros. That belongs in your cost line per order, otherwise your margin on imports no longer adds up from July onwards.

Why a standard package does not solve this for you

Shipping software and standard inventory packages cover the happy path: one webshop, one country, fixed rates. They do not cover your exceptions. Not the combination of bol SLAs with your own carrier contracts, not the split logic for heavy orders, and not writing the actual costs back to margin per supplier. That is where I build the connection that brings your data from shipping software, channels and bookkeeping together, with the calculation rules that fit your operation. That is build-vs-buy, and the exceptions are exactly where the margin sits.