Three months ago we ran a 3PL transition for a client. At the point stock was due to leave the outgoing provider we found over £100,000 of variance between the physical count and the system records. We picked it up before anything moved, because the exit count was built into the plan as a gate rather than a task, and it got investigated and closed out while both providers were still engaged and both had a reason to care.
I've been involved in somewhere north of twenty of these now, between 3PL moves, warehouse relocations and bringing fulfilment in-house. The pattern is consistent enough that I think it's worth writing down. The transitions that go badly rarely go badly for the reasons people plan for.
Most brands approach a warehouse move as a logistics exercise. Dates, transport, racking, systems integration, go-live. That's a reasonable instinct because you are physically moving stock from one building to another.
The problem is that the logistics side is the part your new provider already knows how to do. They run onboardings for a living. They'll give you an implementation manager, a template plan and a set of milestones, and most of that will happen whether you push it or not.
What nobody owns by default is the inventory. Which stock moves, which doesn't, what condition it's in, whether the numbers are right, and how you keep servicing demand while it's in transit. That work sits with the brand, and in my experience it's the workstream most likely to be underestimated.
When stock comes out of an exiting site, two things tend to be true. The condition is worse than anyone expected, and the physical count doesn't match the system.
Damaged outers, mixed pallets, lines that have been sat in a corner for two years that nobody has looked at. On the numbers, variance can be significant, and it's rarely in your favour.
If you don't reconcile at the point of exit, you lose the ability to resolve it at all. Once stock has been transported, received into a new building, put away under new location logic and mixed with new inbound receipts, there is no way to establish where a gap came from. Lost at the old site, damaged in transit, miscounted on receipt, or never actually there. Each of those has a different owner and a different remedy, and after the move you can't distinguish between them.
By that stage you are also negotiating with a provider you have just given notice to. They have no commercial reason left to help you investigate, and depending on how the exit was handled they may have a reason not to.
So we treat the exit count as a control gate. Stock doesn't ship until it has been counted, reconciled against system records, and the variances have been investigated and signed off by both sides. It takes time and it is not an interesting piece of work, but it is the single point in the process where a discrepancy can still be resolved with everyone in the room.
A reconciliation that only counts units misses the other half of the opportunity.
Some of what's in that building shouldn't come with you. Damaged stock, obsolete lines, superseded packaging, product that was never going to sell. Every unit of it costs freight, costs a receipt line at the new site, costs storage, and then costs a write-off further down the line when someone finally makes the decision.
The exit is the cheapest point at which you will ever deal with it. The stock is already being handled and counted, and you have a natural reason to make a call on it. Take the decision then and arrange for the rest to be recycled or destroyed rather than shipped.
This is the part that makes the difference between a transition plan and a date in the calendar.
Going into a move you are working with two sources of supply: what's currently sitting at your outgoing provider, and what's inbound from suppliers. You need to understand both, and you need to understand them against demand.
The questions I'd want answered before committing to a go-live date:
Work through those four and the shape of the transition plan becomes obvious. It might be that new inbound goes direct to the new provider while you run legacy stock down at the old one. It might be a single bulk transfer. More often it's a phased combination, and the phasing is what protects service levels.
I'll be honest about something. Across the transitions we're running at the moment, the stock workstream is consistently the one tracking behind. Tenders progress, contracts get signed, integrations get built, and the stock move plan sits waiting for someone in merchandising or planning to find the time. It's the hardest piece of analysis and the least visible, so it gets deferred, and then it becomes the critical path.
Your incoming 3PL will hand you an onboarding plan built around a single outcome: everything required to pick, pack and dispatch an order. Every line on that list is necessary eventually. But you don't need all of it at the same time, and treating it as one block is a common reason go-live dates move.
I'd break it into four stages and hold the provider to each in turn.
Receive. What is the genuine minimum for them to take a pallet in? SKU set-up, product data, ASN handling, goods-in process. That's a considerably shorter list than the full onboarding scope, and receipts have to start well before anything ships out.
Test receiving properly. Because it's the first thing that goes live, it's the first thing that will expose data problems. Build a testing plan around it rather than assuming it'll work.
Replenish. Getting stock from goods-in to a pick face is a separate problem with its own configuration.
Dispatch. Full order integration, carrier setup, returns flow, reporting.
The same logic applies to your delivery proposition. There's a strong temptation to launch with the full carrier matrix on day one: standard, express, locker, collect, store drop-off, across every market. I'd start with one service per country, get live, prove it, and add options once the operation is stable. Complexity introduced at go-live is complexity you'll be debugging during your first peak.
I've seen brands switch without telling their incumbent, on the basis that there's no contract compelling them to. The provider finds out when the daily order file stops arriving.
I don't think much of it ethically, but set that aside, because it's also a poor commercial decision. You typically have weeks or months of trading left with that provider before the switch. You need them to keep performing through it, you need their cooperation on the exit count, you need their data, and you need a clean handover rather than an obstructive one.
Ninety days is a reasonable standard even where nothing in a contract requires it. You'll get a better exit and a better reconciliation, and you'll be able to pick up the phone in six months if something comes to light that you can't explain.
If you've got a 3PL transition ahead of you, the sequence I'd follow is this. Get the exit reconciliation into the plan as a gate, with stock unable to move until it's counted, reconciled and signed off by both parties. Make the decision at exit on what travels and what gets written off. Build the stock strategy across legacy and inbound against your demand profile, and start it earlier than feels necessary, because it's the workstream that slips. Phase the readiness plan into receive, test, replenish and dispatch. Pick a go-live date where you'll have real support through hypercare. And give proper notice.
None of it is complicated. It's just the work that gets deferred because it's difficult and unglamorous and there's always something more urgent.
We use a standard gate document on every transition. It sets out what has to be counted, what has to be agreed, and who signs before stock is allowed to move, along with a variance log for working through the discrepancies.
Most of this work sits with the brand, and the constraint is usually that the person who can run it is also the person running the current operation. If that's roughly where you are, we get brought in for transitions like that. If you would like to know how we approach a fill, fill out our contact form, and I will jump on a call with you. If you plan to do it yourself im happy to share some of the templates we use with you.
For a single-site UK move, twelve weeks from contract signature to go-live is realistic, assuming your data is in reasonable shape. Multi-territory transitions take longer, and the constraint is usually stock planning and data readiness, not anything the 3PL is doing.
It's a full physical count of stock at the outgoing site, matched against system records, with variances investigated and resolved before anything is transported. It matters because once stock has moved and been received into a new site, it becomes impossible to establish where a discrepancy originated.
No. Use the exit as the point to identify damaged, obsolete and slow-moving stock and deal with it rather than paying to transport, receive and store it
Check your contract first, but ninety days is a reasonable standard even where no notice period is specified. You need that provider's cooperation through the exit, so handling it properly is in your interest as well as theirs.
Any date where your hypercare period falls across a shutdown. Early December is a common trap, because the first few weeks of live operation land in the Christmas and New Year period when support on both sides is thin.
Not always. If you have someone internally with the bandwidth and the operational experience to run it, they can. The problem most brands hit is that the person who could do it is also the person running the current operation, and the transition becomes a second full-time job.
Posted on 13th September 2026