A practical guide to auto-split orders for multi-warehouse UK sellers
What auto-split actually does, the rule types that matter, and how to pilot it without breaking dispatch for a week. With real numbers from UK sellers.
Most dispatch pain in a multi-warehouse UK operation comes from one specific class of order: the one that needs to be split across two or more parcels. Before automation, splitting happens in someone's head in the warehouse, which is where mistakes creep in.
Auto-split is the automation that takes one incoming order and fans it out into the right set of outgoing parcels — before it ever reaches the picker. This guide walks through what it actually does, the rule types that matter, the numbers from real UK operations, and a pilot plan that doesn't break dispatch for a week.
The problem, in one example
A customer orders three items in one transaction:
- A 300g ceramic mug — stored in Warehouse Milton Keynes.
- A 2kg ceramic plant pot — stored in Warehouse Birmingham.
- A 150ml perfume — stored in Milton Keynes, but flagged hazmat.
No single courier and no single parcel solves this. The order has to ship as three separate shipments:
- Mug alone — small parcel, Royal Mail Tracked 24.
- Plant pot alone — medium parcel, Evri or DPD.
- Perfume alone — hazmat-approved courier (not everyone will take fragrances).
Without auto-split, that's three manual decisions for every order like this: warehouse assignment, parcel sizing, courier selection. At 1% of your volume, it's a nuisance. At 5%, it's a slow daily drag. At 15% (common in homeware or mixed-category sellers), it dominates your dispatch manager's day.
The five reasons orders actually need splitting
These are in roughly descending order of how often they come up:
- Multi-warehouse. Items in physically different warehouses can't share a box. Hard constraint.
- Hazmat and restricted items. Perfume, lithium batteries, aerosols, magnets, knives, alcohol. Each has courier-network restrictions and usually can't share parcels with arbitrary items.
- Courier weight and size thresholds. Royal Mail Tracked 24 caps at 20kg; some orders push over. Evri drops price dramatically below 2kg; splitting a 2.2kg order into 1.5kg + 0.7kg can save 35-50p.
- Perishable or time-sensitive SKUs. Food items, flowers, live plants — these can't wait on a slower item.
- Age-restricted items. Alcohol, tobacco, CBD — need signature-on-delivery premiums that the rest of the order shouldn't pay.
What a good auto-split actually does
Every platform that claims auto-split handles warehouse-level splits. The real differentiators are:
- Attribute-based splitting. Tag SKUs as hazmat, fragile, perishable, age-restricted. Let rules fire based on any combination of tags.
- Weight/dimension rules. "If combined weight > 10kg, split into the largest two sets ≤ 10kg each."
- Courier-rule splitting. "If any item requires Royal Mail Tracked but another requires DPD Next Day, split into two shipments, each using its correct courier."
- Manual override in the dispatch queue. For edge cases — and there are always edge cases.
- Child-order rate shopping. Each child shipment picks its own cheapest courier, independent of the parent.
The numbers (from real MaxInvent customers)
Data from UK sellers on MaxInvent in Q1 2026:
| Seller profile | Multi-item order % | Orders needing split | Time saved /month |
|---|---|---|---|
| Homeware, 2 warehouses, 3,000 orders/mo | 32% | ~160 | ~5.3 hours |
| Beauty, 1 warehouse, hazmat SKUs, 5,000 orders/mo | 18% | ~95 | ~3.2 hours |
| Electronics, 3 warehouses, 8,000 orders/mo | 25% | ~450 | ~15 hours |
| Wholesale/B2B + retail, 2 warehouses, 1,200 orders/mo | 47% | ~280 | ~9.3 hours |
At £15-18/hour loaded labour, this is £50-£270/month saved purely on admin time. More importantly, it eliminates a whole class of dispatch error — the mis-picked or mis-labelled split order — each of which costs £8-15 in handling plus the opportunity cost of an unhappy customer.
How to pilot auto-split without breaking dispatch
A safe rollout looks like this, over roughly four weeks:
Week 1: Tag your SKUs
Before any automation, you need clean data. Pick the top 10-20% of SKUs by volume and flag them with the attributes that matter: warehouse, weight, hazmat, perishable, age-restricted, oversized. Most good inventory platforms have bulk-tag tools so this is 1-2 hours of work. Don't try to tag all SKUs upfront — you'll lose momentum and quit.
Week 2: Run in "suggest" mode
Turn on auto-split for those tagged SKUs in suggest-only mode. The platform proposes splits on incoming orders but doesn't execute. Every morning, review the proposals against what your dispatch team would have done manually. Tune the rules where they disagree.
This is the phase where you catch rule gaps. Example we saw at a beauty seller: the platform proposed splitting a lip-balm-plus-perfume order into two parcels on hazmat grounds, but the lip balm was under the threshold for hazmat classification — the rule needed refining.
Week 3: Flip to execute for piloted SKUs
Once the suggestions match what you'd do manually 95%+ of the time, flip to execute mode for those same tagged SKUs. Monitor the dispatch queue for the first 3-5 days. Keep the manual override button visible in case.
Week 4: Roll out to all SKUs
If week 3 went cleanly, expand to the full catalogue. Tag remaining SKUs as a background task — use order-triggered tagging (when a SKU appears in an order for the first time, prompt for attributes) so you fill in the tail over time instead of upfront.
What to check when evaluating platforms
If you're comparing inventory platforms and auto-split matters to you, ask each vendor:
- "What rule types can I configure? Warehouse only, or attribute + weight + courier + custom expression?"
- "Can I override a specific order's split in the dispatch queue before it picks?"
- "Does each child order rate-shop its own courier, or does the whole order inherit one?"
- "What happens when one child shipment can't be fulfilled — does it block the others?"
- "Show me the audit log for a specific split decision. Can I tell which rule fired?"
Platforms that answer 1-2 convincingly are credible. Platforms that can't answer 5 have opaque logic — when something goes wrong you won't be able to diagnose it.
The short version
Auto-split is a must-have the moment you add a second warehouse, or the moment you have more than 5% of orders needing multi-carrier split. The cost is low (most inventory platforms that include auto-split bundle it in the base tier). The payoff is hours per week back and fewer returns.
The hard part is not the software. It's the data hygiene — SKU tagging, courier rule definition, team training. Do the pilot carefully and it becomes invisible infrastructure. Skip the pilot and it's how you accidentally ship a hazmat item with a regular courier.
If you're evaluating platforms on auto-split specifically, our answer page has a platform comparison with the feature-by-feature breakdown.