-
Pallets fit the truck, not the rackingThe pallet dimensions decide the truck, and no dock decides it again. Both can shrink the load you thought you booked.
-
A kit is not a unitSets get opened for parts, so the count you are planning the move against can be a fiction.
-
Every order source, mappedA retail or wholesale order can ship out of the same building in the middle of the move, on a deadline that does not care about your transfer.
-
One cutover lineMove in waves and keep the old 3PL fulfilling, so nothing ships twice and nothing falls through the gap.
The plan is clean on paper. The move happens one level down from it.
The plan for moving to a new 3PL always reads clean. Sign, transfer inventory, connect the software, go live in a week or two. I wrote the full arc of that process in choosing and working with a 3PL, and Phase six is the inventory transfer and software setup. This note lives one level below that phase. It is the set of details that do not make the plan, the ones that surface only once the trucks are booked and the inventory is in motion, pulled from a recent move I ran between two providers.
None of these are exotic. They are the difference between a transfer that goes the way you drew it up and one that costs an extra week and a few thousand dollars you did not budget. The pattern is always the same: the assumption that felt safe on paper turns out to have a physical or a data reality behind it that nobody checked.
Pallets fit the truck, not the racking
The freight side of Phase six is where you confirm pallet count, weight, and dimensions before anyone books a truck. Two things underneath that confirmation quietly decide what you can actually book. The first is the dock. On the recent move, neither of the outgoing buildings had a loading dock, and a 53-foot trailer with a lift gate is close to impossible to source, so the 53-footer was off the table before we discussed a single date. The second is pallet size. Once we were down to smaller box trucks, 13 pallets at their built dimensions would not fit on a 26-foot truck, so we cut the pallet count to make the load fit and ended up running two small trucks of about a dozen pallets each instead of the one big truck the plan imagined.
The lesson is that the pallet dimensions decide the truck, and the dock decides it again, and neither shows up on the tidy version of the plan. A pallet built to fit the outgoing warehouse's racking is not automatically a pallet that fits the truck you assumed. Confirm dock, lift gate, and real pallet dimensions on both ends before you promise anyone a pickup date, because a wrong assumption here does not cost a phone call, it costs a truck.
A kit is not a unit
This is the one that quietly breaks the count. If you sell sets, the number in your system is often a fiction, because a kit gets opened for its parts. On the move, the outgoing inventory was a mix of cartonized units, palletized units, and loose boxes of broken-open parts, and the count the outgoing 3PL gave us was not one we could trust to plan against. So we did not try to reconcile it in their building. We moved everything and did a single clean count on arrival at the new warehouse, which is both more accurate and cheaper than paying two teams to count the same messy shelves twice.
The takeaway is to find out how kits and bundles are actually stored and counted before you build a plan on a unit number. A transfer planned against a count that does not survive contact with the shelf is a transfer that arrives wrong, and you find out at the worst possible time, after the truck has already left.
Every order source, mapped
The software setup in Phase six is easy to picture as one integration: connect the store, orders flow, done. The reality is that a brand at any scale has more than one place orders come from, and they do not all run through the hub you are wiring up. In the middle of this transfer, a big-box retail order of 111 cartons had to ship out of the same building on a hard void date, on a Monday pickup, while the DTC inventory was being staged to leave. That order had its own truck, its own deadline, and its own claim on the same pallets and the same floor space, and none of it was in the go-live plan.
So map every order source before you cut over, name where each one lands, and check the ones that do not flow through the main integration. The retail and wholesale orders with hard pickup windows and void dates are exactly the ones you cannot afford to discover late, because missing one is a chargeback or a lost account, not a delayed DTC package.
One cutover line
Every detail above rolls up to the last one. There has to be a line, agreed by both sides, that separates what the old 3PL still ships from what the new one owns. The way you draw it in practice is to move in waves rather than all at once, and to keep the outgoing warehouse fulfilling until the new one is genuinely live, the same discipline as running a test phase on a small slice of live orders before you scale. Without that line you get the two failure modes that define a bad transfer: an order ships twice because both buildings thought it was theirs, or it ships from neither because both thought it was the other's.
The party most likely to blur the line is the one you are leaving. On this move the outgoing 3PL was two people with limited staging space and slow replies, which is common, because the outgoing provider is the least motivated party in the whole transfer and often not even on the same tools you are. Get a direct line into that building early, decide who owns communication to each party, and confirm the cutover line is one they are holding too, not just one you wrote down. This is the same reason to lock in the operational basics the week before you go live.
That last point is the whole note in one sentence. A transfer is not one company handing off to another, it is two operations running at once for a stretch of days, and the gap between them is where orders go missing. The plan says go live. The move is everything you did to make sure that day had a clean edge.