Headquarters
175 S Main St Suite 1310,
Salt Lake City, UT 84111
A team spends a quarter reducing split shipments. The split rate falls, but shipping cost per order keeps climbing.
The result feels contradictory only because the metric and the cost are measuring different things. Split rate tells you how often an order leaves in more than one parcel. It does not tell you what was inside each parcel, how much margin it carried, or whether the extra shipment was worth creating.
A $200 jacket shipped separately may protect a high-value sale and meet the customer’s delivery promise. A $12 accessory sent in its own box can consume most of its value in fulfillment and transportation costs.
Both count as one split. Their economics are completely different.
Split shipments are sometimes necessary. Inventory is distributed across stores and warehouses, sizes sell unevenly, and no single location may hold the complete order. Splitting can prevent a cancellation or keep an expedited order on time.
The cost problem begins when each resulting parcel repeats work that the order already paid for once. Every additional shipment has its own pick, pack, packaging, label, carrier charge, and delivery journey.
Capgemini reports that last-mile delivery accounts for 41% of total logistics supply chain costs. That makes the final leg the largest cost component in many fulfillment models. Sending another parcel does not add a small administrative fee. It creates another shipment that has to travel through the expensive part of the network.
The contents of the parcel determine whether that choice makes financial sense. A high-value item may carry enough margin to justify separate fulfillment. A low-value remainder may not.
That is why “never split” is not a useful policy either. Avoiding every split can delay high-value items, weaken delivery promises, and hold revenue while the network waits for one facility to accumulate the complete order.
The goal is not the fewest splits at any cost. It is fewer uneconomic splits.
Split rate is popular because it is simple:
Number of orders fulfilled in multiple shipments ÷ total fulfilled orders
The calculation is useful as a network signal, but it is incomplete as a margin metric.
Consider two retailers with the same 15% split rate. The first mostly splits higher-value orders when no location can fulfill the complete basket. Each parcel carries enough value to support its fulfillment cost, and the split protects delivery speed.
The second retailer splits accessories, promotional items, and other low-value remainders into standalone parcels. Its rate looks identical, but its margin outcome is worse.
A team that manages only the percentage may spend months reducing profitable splits while leaving the costly ones untouched.
Standard routing tools can reinforce this focus. Shopify’s Minimize split fulfillments rule prioritizes locations that can fulfill an order in the fewest packages. That is useful consolidation logic, but Shopify’s standard routing-rule list does not include a native rule that evaluates a proposed shipment against a minimum value.
Package count and shipment value answer different questions:
Package count asks, “Can we fulfill this order in fewer boxes?”
Shipment value asks, “If we create another box, is what we are sending worth the added cost?”
A strong routing strategy needs both.
Using total order value alone can still hide the problem.
A $250 order may look profitable enough to split. But if $238 of merchandise ships from one facility and a $12 accessory ships from another, the total order value says nothing about whether the second parcel makes sense.
The economic test belongs at the proposed-shipment level. Before creating a split, routing should evaluate both sides:
Checking only the first group can leave a low-value remainder behind. Checking only the remainder can approve an equally weak outbound parcel. Both proposed shipments need to clear the retailer’s minimum value.
This does not mean merchandise value is a perfect substitute for contribution margin. A useful threshold should reflect the retailer’s actual economics, including typical gross margin, pick-and-pack expense, packaging, parcel transportation, expected surcharges, and the value of meeting the delivery promise.
The threshold is a practical guardrail. It prevents the routing engine from treating a $12 accessory and a $200 jacket as equally reasonable standalone shipments.
An arbitrary threshold can create a different problem. Set it too low, and low-value parcels continue to leave the network. Set it too high, and orders wait even when a split would protect a valuable sale or an urgent promise.
Start with recent split-order data. For each parcel, compare merchandise value and estimated contribution margin with its incremental fulfillment cost. Then look for the point where a standalone shipment regularly stops covering that cost.
The analysis should also account for the consequence of not splitting. An expedited order, a VIP commitment, or an item with a short promise window may justify an exception. A standard order with a longer delivery window may have more time to wait for consolidation or another routing attempt.
The Shipment Threshold should express a policy the business can defend:
“We create a separate shipment when each resulting parcel carries enough value to justify its incremental cost.”
That is much clearer than targeting a split rate without knowing which splits the target will remove.

HotWax Commerce adds shipment value to the brokering decision instead of reviewing the cost after fulfillment.
Retailers configure a Brokering Shipment Threshold for the product store and use the Shipment Threshold Value Check within the relevant routing inventory rule. When an allocation would split an order, the brokering engine checks the value of both the items it proposes to allocate and the items that would remain unallocated.
If both sides meet the threshold, the split can proceed through the rest of the routing logic. If either side falls below the threshold, HotWax does not create that low-value split through the rule. The unallocated items can remain available for a later routing path rather than automatically becoming an uneconomic standalone parcel.
The threshold works alongside the rest of configurable order routing. Facility groups define which stores or warehouses can participate. Proximity, inventory availability, safety stock, and facility order limits narrow the eligible locations. Partial-allocation settings determine whether the route may split an order, while the shipment threshold determines whether the proposed split carries enough value.
This distinction matters. Partial allocation answers whether splitting is permitted. The shipment threshold answers whether this particular split is financially acceptable.

Because the decision happens during brokering, the rule is applied before inventory is committed to separate fulfillment locations and before a label is created. The operations team is not left reviewing a costly parcel after it has already entered fulfillment.
Split rate can remain on the dashboard, but it should sit beside measures that explain the economics:
Average merchandise value per parcel on split orders.
Number and value of standalone parcels below the chosen floor.
Estimated incremental fulfillment cost from split shipments.
Contribution margin after fulfillment by split and non-split orders.
Age and promise performance of items held for another routing attempt.
These measures show whether the policy is protecting margin without creating a new delivery problem. A lower count of low-value parcels is useful only if held items are still resolved within the customer promise.
They also make threshold reviews more productive. Instead of debating whether a 15% split rate is too high, the team can see which parcels are paying for themselves, which are not, and what happened when low-value remainders were held.
Split rate measures frequency, not financial quality.
An extra parcel repeats fulfillment and delivery costs, so its contents need to justify that expense.
Total order value is not enough. Evaluate both the proposed shipment and the remainder.
Set the minimum value from contribution margin, parcel cost, and delivery commitments, not an arbitrary percentage.
Use shipment value with consolidation, inventory, eligibility, capacity, and promise rules rather than as a standalone policy.
* * *
Book a demo to see how shipment-value threshold would protect margin without applying a blanket ban to every split.