Home
Post Purchase Operations

Split shipments: two boxes, one promise

VerifiedVerified & Reviewed
A split shipment is one order delivered in multiple parcels. Split when waiting would break the promise, never for warehouse convenience, tell the customer at dispatch, and watch each line item, because a half-delivered order is where promises quietly break.

A split shipment is one order delivered in more than one parcel: items in different warehouses, different suppliers, or different availability dates, so the order ships as it can rather than as one box. Operationally it is sometimes the right call. My frame for it, from two decades running ecommerce brands: the customer did not order parcels, they ordered an outcome, and however many boxes you use, there is still exactly one promise. Every order is a promise. Keeyu keeps it, across however many shipments it takes. The mistake is not splitting orders. The mistake is letting the promise split with them, which is exactly the visibility gap proactive e-commerce operations, the system of action for e-commerce, exists to close.

When splitting is right, and what it costs

Split when waiting would break the promise: half the order can ship today and the other half is a week out, so the first half goes. The order management system page covers where that routing decision should live. Do not split for warehouse convenience, because every extra parcel is a separate carrier charge, separate packaging, and separate labor, and two boxes can nearly double one order's shipping cost. Our own operating rule with brands is blunt: split only when you are genuinely sending two different packages, an item on pre-sale or out of stock, the free protein shaker to follow, send what you have and the rest when it is available. The routing logic should always try the consolidation question first: is there one location that can fill the whole order inside the promise, even at slightly higher cost? The complexity scales with the catalog. One brand we serve processes only 800 to 1,000 orders a month, but at a $15,000 average order value, coordinating multiple suppliers per order is the entire operational problem. Another, a nutrition business doing about 600,000 orders a year, runs pre-orders, bundle deals, and subscription bundles, where splitting is structural rather than occasional. The everyday version is smaller and just as messy: a customer orders an air fryer and a hair curler that ship from two different suppliers, and one order becomes two half-truths in your systems.

One order card splitting into two parcels on separate routes, with a single promise bar arcing over both and reconnecting at the doorstep.

Where split shipments silently break

The failure modes cluster in the gaps between systems, which is why they go unseen. The order shows part-fulfilled and stops being anyone's job, a state covered in depth on the order fulfilled page. One parcel delivers and the other stalls, and your platform's single status field cannot say that. Refunds get ugly: we found one brand whose bundled orders could not be partially refunded or handled at the line-item level at all, a structural limitation nobody had noticed because no system was watching orders at that grain. And the support queue tells the story last: slow freight, split shipments, and no consolidated visibility across systems is the exact triad I see on discovery calls, and by the time it is visible in tickets, the customer has already had the bad version of the experience. Fulfillment breaks silently. Something has to hear it first.

Splits also drive returns: an unbundled order arriving in pieces gets sent back in pieces, which the return rate formula page is the right way to measure.

Two boxes, one conversation

The FTC's Mail, Internet, or Telephone Order Merchandise Rule treats each unshipped item as its own obligation with its own clock, which is a useful way to think about splits operationally too. Every split is a communication event. At dispatch, the customer needs to know their order is arriving as two deliveries, what is in each, and when each lands, through the milestone flow on the shipping notifications page. Miss that and you have manufactured a guaranteed "where is the rest of my order" ticket at double the investigation cost, because now the answer spans two carrier journeys. The deeper fix is watching each line item, not each order, against the promise: Keeyu tracks every order at the grain where splits actually live, across storefront, warehouse, suppliers, and carriers, catches the half that stalls, and acts before the customer writes in. Detect. Decide. Act. Keeyu gets shoppers what they want, on time, as promised, in however many boxes. The architecture view is on the order fulfillment system page, the discipline on the post-purchase operations hub, and a demo will show your own split-order blind spots. Every order is a promise. Keeyu keeps it.

Frequently Asked Questions

What is a split shipment?

One order sent in more than one parcel, usually because items sit in different warehouses, ship from different suppliers, or become available at different times. The customer placed one order and now has two or more deliveries to receive.

Why do split shipments cost more?

Every extra parcel is a separate carrier charge, separate packaging, and separate handling. Two boxes for one order can nearly double the shipping cost of that order, which is why smart routing tries to consolidate before it splits.

Should I tell customers when their order ships in multiple packages?

Always, and before they notice. One order arriving as two deliveries on different days generates a "where is the rest of my order" ticket unless each parcel's contents and dates are communicated clearly at dispatch.

How do I reduce split shipments?

Route orders to the location that can fill the whole order first, even at slightly higher single-parcel cost. Split only when the delay from waiting exceeds the promise, and treat every split as a communication event, not just a fulfillment event.

References

No items found.

Ready to Stop Reacting?

The fastest way to see how Keeyu prevents complaints is to see it in action.

In one call, we’ll map your current operations, show how our AI Agent fits in, and walk through real examples of issues fixed before customers notice.

Most teams go live within 48 hours. We never share your data.