EU Orders To EU Customers
Stopped Costing More To Fulfil
Replacement of a single GB fulfilment origin serving all EU demand with a two-node UK and Netherlands model, selected through a vendor-neutral 3PL RFP and integrated across the order layer, storefront and returns platform.
80
%
EU Orders Covered By The EU Node
1
New EU Fulfilment Partner
From 4 → 0
Customs-Controlled Movements Per Returned And Resold EU Pair
Client Profile
D2C Footwear Brand
~£24m net revenue · ~25k orders/m shipped, ~38k at peak.
Average Basket Value ~£115, with returns at ~30%.
Single UK 3PL fulfilling all demand from one GB site.
1/3 of orders ship to EU markets, led by Germany, France and the Netherlands.
Product manufactured in Portugal and Italy.
The Challenge
Every order was picked and despatched from one GB site. Stock manufactured in Portugal and Italy entered GB under the applicable preferential treatment where the required origin evidence was available. It was then exported back to EU customers as an individual cross-border consignment, with an import declaration on every order. Every EU return followed the same route in reverse, back to GB for inspection and restock, before being exported again if it was resold to another EU customer.
From 1 July 2026, the EU replaced the €150 duty-free threshold with a temporary €3 charge per tariff-classified item on qualifying low-value imports. Higher-value orders moved into normal tariff treatment, while the temporary charge is not automatically repaid on returns and is expected to be replaced by standard tariff treatment from July 2028.
The business was therefore moving EU-manufactured product out of the EU and importing it back into the same market, creating repeated customs movements and duty exposure across approximately a third of its volume. The cost was due to increase under the published regulatory timetable, but the routing itself was a legacy of where the business had grown from rather than a deliberate supply-chain decision.
Why EQ50 DIGITAL
A node decision falls across three kinds of supplier.
A broker rules on origin and classification but does not select warehouses. A consultancy builds the business case but does not deliver the integration. A systems integrator connects a second node without holding a position on whether it should exist.
EQ50 DIGITAL was appointed across all four stages: modelling, selection, integration and cutover.
Those stages are dependent. The exposure had to be sized at line level before the investment could be justified, the node tested against lighter alternatives before capital was committed, and the provider selected on operating fit before anything was built. A single appointment kept the reasoning intact from the model through to go-live.
EQ50 DIGITAL owns the business case, the provider selection, the integration and the technological readiness required to bring the node live.
The EQ50 DIGITAL Approach
01
Clarify the business priorities
A session with Finance, Operations and Ecommerce set the order: remove the structural duty cost on EU volume, protect UK service through the change, and avoid two large operational changes in one year.
The third priority shaped the programme. The UK 3PL contract was near review and was held back deliberately, so an EU node was not built on top of a UK transition.
The customer-facing returns proposition – policy, pricing and available methods- was raised and deferred on the same logic. The operational changes required to route EU returns to the new node, inspect them and return eligible stock to sale remained inside the programme. A later phase would address how the customer initiated and paid for the return.
02
Diagnose and scope
Lighter options were tested first and rejected on stated grounds. Better cross-border parcel rates and consolidated linehaul reduced freight cost but left the import declaration and most of the customs exposure in place under the existing route. A European returns hub improved the reverse leg but did not address the outbound movement.
Direct factory fulfilment was also tested. It was rejected because the factories could not support the required single-order picking, inventory visibility, carrier coverage and consumer-returns model at an acceptable cost, rather than because direct fulfilment was technically impossible.
Size-curve modelling established that a node holding core styles and core sizes could serve around four fifths of EU demand, which set the allocation and kept the GB to EU lane open for the tail.
03
Design the solution
An RFP went to five EU providers and a shortlist of three, scored on footwear handling and size-level pick accuracy, returns inspection and in-market restock, pooled carrier depth across the target markets, integration maturity against the order layer, peak capacity, and three-year cost including inbound handling, storage and per-order fees.
The Netherlands was selected over lower-cost central European options using the weighted location of demand and returns, inbound routes from Portugal and Italy, carrier coverage across the target markets, service levels and total operating cost. The decision accepted a higher fulfilment cost per order in exchange for stronger coverage of the largest EU markets.
Routing rules were written to serve an order whole from one node wherever possible, splitting only where a tail size or stock constraint required it. Factory inbound was redirected so the EU allocation moved directly into the Netherlands node and did not enter GB.
04
Deliver the change
Integration Readiness came first: what the storefront exposed for inventory by location and market-specific delivery options; what the order layer could support for whole-order routing, stock fallback and split orders; and whether the returns platform could operate a second return location.
The discovery confirmed which existing native capabilities could be retained and where direct integration was required. Inventory and routing used the capabilities already available in the order layer; a direct API carried despatch and tracking events from the 3PL and the returns platform was configured with the new return location.
Built to that pattern: available-to-sell inventory by node, safety stock, whole-order routing, defined split-order handling, and despatch and tracking events feeding one customer-facing journey regardless of which node served the order. Delivery options and dates were configured by market and serving node.
The exception design covered failed inventory updates, node rejection, partial despatch, GB fallback and customer communication where an order could not be fulfilled as originally routed.
In parallel, the fiscal workstream established the legal owner of the EU stock, the Dutch VAT registration, Union OSS reporting and the transaction data required from the 3PL. The treatment was confirmed by the client’s appointed tax adviser and translated into the finance, order and reporting requirements for go-live.
Cutover ran market by market in the two months after peak.
Key Decisions & Trade-Offs
The node was justified on structural customs cost, not on delivery speed.
Faster EU delivery was the easiest case to make internally and the weakest basis for the investment, because the existing cross-border service was acceptable to customers. The figure that carried the decision combined repeat customs declarations, temporary €3 exposure, applicable tariff exposure on higher-value orders and the expected post-2028 treatment, using assumptions confirmed by the customs broker.
That ties the business case to a regulatory and customs position that can change. The model was therefore tested against delayed implementation, lower-duty and relief scenarios. A delay to the expected increase slowed the payback but did not remove the underlying cost of moving EU stock through GB before selling it back into the EU.
The EU carrier contracts stayed with the 3PL.
The brand had no EU shipping history and no in-market volume to negotiate against. At this scale the provider’s pooled volume beat anything achievable directly, so carrier depth per market was scored inside the RFP rather than pulled out as a separate negotiation.
The cost is ownership. The brand does not hold those contracts, the rate detail or the direct carrier relationship, and sees the provider’s view of performance so contractual reporting, carrier substitution rights and access to service-level data were therefore written into the 3PL agreement, even though the individual carrier rates and direct relationships remained with the provider.
Core sizes to the node, the tail from GB.
Core styles and sizes covered around four fifths of forecast EU demand using twelve months of order data. The remaining fifth – approximately 20,000 parcels a year at the existing volume – continued to ship cross-border from GB, so customs-controlled movements and the associated exposure were not removed from the whole book. A smaller share of orders also split across two nodes where one location could not serve the basket whole.
The alternative was a materially larger inventory commitment made before EU-node sell-through, replenishment behaviour and the level of stranded stock had been proven. The allocation was therefore designed to widen after launch rather than duplicating the full range from day one.
The Outcomes
~80%
Modelled share of EU orders covered by the core styles and sizes allocated to the node at go-live.
4 → 0
Customs-controlled movements in the lifecycle of an EU-node pair that is sold to an EU customer, returned and resold. Stock moves from the EU factory into the EU node, ships within the EU, returns within the EU and is resold from the same customs territory.
−2 days
Median order-to-delivery time for EU-node orders in the first full quarter after go-live, measured against the equivalent markets and services in the pre-engagement cross-border baseline.
Intra-EU
Orders fulfilled from EU-node inventory move through intra-EU parcel networks without a consumer-order import declaration or temporary €3 customs duty. This applies to inventory held in the EU node, the GB served tail retains its existing customs treatment.
Second Fulfilment Node
An additional fulfilment location and defined fallback route where there had previously been one. Resilience applies where the alternative node holds the inventory and is configured to serve the affected market.
The Handover
The client owns the provider relationship and performance framework, the customs-broker and tax-adviser relationships, the allocation model that decides what sits in each node, and the routing rules that determine which node serves an order.
Handed over with it: an allocation review tied to sell-through and stock cover, so the EU range can widen without re-engagement; integration documentation covering inventory, routing, despatch, tracking and returns across both nodes; a fallback and exception runbook; and the modelled customs position alongside the broker’s classification, origin and relief assumptions.
The handover also includes the fiscal data map agreed with the client’s tax adviser, covering Dutch VAT, Union OSS and the transaction information required from the 3PL, so the client can maintain the reporting model as markets, volumes and inventory allocation change.