Zero-Downtime WMS Implementation: How UK Distributors Go Live Without Stopping Orders (2026 Guide).
Published: 01/10/2026 | Last updated: 01/10/2026 | Reading time: 12 minutes
Quick answers
What is a zero-downtime WMS implementation?
A warehouse management system go-live planned so the warehouse is never without a working system. The old system or process remains available until the new WMS is proven, using parallel running or a phased rollout plus a tested rollback plan.
How long does a WMS implementation take in the UK?
A single-site cloud deployment for a UK distributor typically takes 8 to 16 weeks from kick-off to go-live. Multi-site, heavily customised or integration-heavy projects take 4 to 12 months.
Can you implement a WMS without stopping operations?
Yes. The combination that makes it work is clean data, two rehearsal migrations, ERP integration tested at full daily volume, a parallel or phased cutover, trained floor champions and a rollback plan with named trigger criteria.
Biggest risk?
Data quality. Not software. Gartner puts the share of WMS implementations that fully succeed at around 24%, and the gap is almost always on the client side: data, integration ownership, testing and change management.
Choosing a warehouse management system is one of the most consequential technology decisions a distribution business makes. It touches every pick, every despatch and every customer promise. The fear that go-live will halt the warehouse is the single biggest reason projects stall at the approval stage.
Downtime is not inevitable. This guide sets out the cutover method Balloon One uses with UK distributors, 3PLs and manufacturers so orders keep flowing while the system changes underneath them. It is written for operations directors, finance directors and IT leads who need confidence before committing budget.
Why WMS go-lives disrupt operations
Most WMS implementations do not fail because of the software. They fail in the gaps between operations, IT and people.
Gartner research cited by Alpine Supply Chain Solutions puts the full success rate of WMS implementations at 24%.
A separate Gartner survey of 306 logistics leaders found 76% of logistics transformation projects miss budget, timeline or KPI targets, and that internal change resistance obstructs projects more than external pressure.
Zebra Technologies found 55% of warehouse leaders say their WMS is not used to its full potential and 40% still rely on manual workarounds after go-live.
Balloon One’s Out of Date research with more than 500 UK supply chain managers found the average warehouse or inventory system was last upgraded nearly six years ago, and over 60% of UK food and drink supply chains still run wholly or partly on spreadsheets or manual processes. The same study put the cost of avoidable internal waste at £156,600 per business per year.
The five failure points, in the order they usually appear:
Poor data quality. Inaccurate SKU masters, locations or stock counts are amplified by the new system on day one.
Weak integration planning. A WMS that cannot reliably exchange orders and stock adjustments with the ERP forces manual workarounds.
Insufficient testing. A handful of test orders in a quiet demo environment is not a full day’s volume.
Unclear ownership. When operations, IT and the board each assume someone else owns the cutover plan, tasks slip.
Resistance to change. Untrained or unconsulted teams revert to old habits the moment pressure builds.
The rest of this guide addresses each in sequence.
How long does a WMS implementation take?
For a single-site UK distributor on a cloud WMS, plan for 8 to 16 weeks from kick-off to go-live. Multi-site, heavily customised or bespoke-ERP projects run 4 to 12 months. The variable is rarely the software. It is data readiness, integration complexity and how much time your team can give the project alongside the day job.
Factor | Shorter timeline (8 to 12 weeks) | Longer timeline (4 to 12 months) |
|---|---|---|
Sites | Single warehouse | Multi-warehouse or multi-country |
ERP integration | Standard Sage or Dynamics connector | Bespoke or legacy ERP with custom APIs |
Data quality | Clean, consistent SKU and location data | Incomplete, duplicated or paper records |
Customisation | Out-of-the-box workflows | Heavily tailored pick, pack or compliance rules |
Internal capacity | Named project lead at 50%+ of their time | Project run from the side of a desk |
Timing. Avoid a Q4 go-live (October to December) unless you have no choice. January to March and May to July are the usual windows for UK distributors: volumes are predictable and the team has bandwidth. Work backwards from the target date and allow at least 12 weeks.
For a fuller breakdown see Balloon One’s WMS implementation timeline guide.
Step 1: Audit how the warehouse actually works
Every zero-downtime go-live starts with a current-state audit. Document how the warehouse runs at 3pm on a busy Wednesday, not how the SOP says it should run.
Walk every process end to end:
Goods-in: How are deliveries booked, checked and put away? Are POs matched at the dock or later at a desk?
Storage and replenishment: Fixed or dynamic locations? How do pickers know where stock is?
Picking and packing: By order, wave or product? How are multi-line orders consolidated?
Despatch: How are carrier labels produced? Who confirms shipment to the ERP?
Returns and adjustments: How are damages, returns and write-offs recorded?
Note every point where paper, spreadsheets or workarounds fill a gap. Involve supervisors and operators; they know the exceptions management does not see, and people support what they help design.
Step 2: Integrate with Sage or Dynamics before you touch stock
For UK mid-market distributors the ERP is almost always Sage (200, Intacct or X3) or Microsoft Dynamics 365 (Business Central or Finance & Supply Chain Management). Integration quality decides whether go-live feels seamless or chaotic.
At minimum the integration must handle:
Sales orders from ERP to WMS for picking and despatch
Purchase orders to WMS for goods-in
Stock levels and adjustments back to the ERP in near real time
Despatch confirmations and tracking numbers to the ERP for invoicing and customer notifications
Sage 200 typically integrates through middleware or a direct database connector. Business Central uses API-based integration, which is flexible but demands careful field mapping, especially unit-of-measure conversions, location codes and batch or serial structures. Field-level mismatches here are the most common cutover-day surprise.
Resolve before go-live:
Sync frequency: real time, every five minutes or hourly?
Failure handling: is there a retry queue, and who monitors it?
Field gaps: which ERP fields have no WMS equivalent, and vice versa?
Test at load. If you process 500 order lines a day, your integration test processes 500 or more.
Step 3: Migrate data without disruption
Data migration is where implementations go wrong quietly. A miscounted location or a missing lot code will not crash the system. It will send the wrong item or leave the oldest stock at the back while fresher product ships.
What to migrate and clean
SKU master: every product, variant and unit of measure. Standardise descriptions, remove duplicates, confirm barcodes match physical labels.
Location master: every bin, shelf, bay and staging area. Replace informal names (“Dave’s shelf”, “the cage”) with a consistent convention.
Opening stock: a full physical count reconciled to the ERP, as close to cutover as possible.
Supplier and customer records: only where the WMS needs them for routing or compliance.
Lot codes, expiry dates and FEFO/FIFO
For food, drink, pharmaceutical and cosmetics distributors, traceability data is not optional. Every lot (batch) code and its expiry date must carry across so that FEFO (First Expiry, First Out) or FIFO (First In, First Out) rules work from the first pick. FEFO picks the stock closest to expiry first and is the default for perishables; FIFO picks the oldest received stock and suits uniform shelf lives.
During migration:
Extract lot codes and expiry dates from the current system or spreadsheets.
Validate every lot code maps to the correct SKU and location.
Standardise date format to DD/MM/YYYY. Mixed US and UK formats are a frequent cause of expired stock being marked pickable.
Quarantine or write off stock already past or near expiry before it enters the new system.
Balloon One’s Out of Date research found 7.29% of perishable stock in UK food supply chains is lost to internal errors such as poor rotation and handling. Clean lot and expiry data at migration is the cheapest fix available.
If you export food to the United States, the FDA’s FSMA Rule 204 requires traceability lot codes at each critical tracking event. UK-only operations will normally meet their obligations with standard lot and expiry tracking. For sector detail see the food and beverage WMS guide.
Treat migration as a rehearsal
Run at least two full trial migrations into a test environment before the real one.
Migration | When | Purpose |
|---|---|---|
First trial | 6 to 8 weeks before go-live | Load all master data and a stock snapshot. Find mapping errors, missing fields, format mismatches. |
Second trial | 2 to 3 weeks before go-live | Repeat with refreshed data. Confirm fixes hold and integration messages flow. |
Final migration | Cutover window | Load production data, reconcile stock, confirm WMS and ERP agree before the first live transaction. |
Each trial gets a sign-off checklist covering SKU counts, location counts, lot code completeness and FEFO/FIFO validation.
Step 4: Choose a cutover strategy
The cutover strategy has the single biggest effect on whether go-live disrupts operations.
Strategy | How it works | Best for | Disruption risk |
|---|---|---|---|
Parallel running | Old and new systems run together for a defined period (typically 1 to 4 weeks). Every transaction is processed in both and discrepancies are investigated daily. | Risk-averse businesses with staff capacity to double-process | Lowest, but high workload during the parallel window |
Phased rollout | WMS goes live one process or area at a time: goods-in and putaway first, then picking, then despatch. Each phase has its own test window and rollback trigger. | Mid-sized or multi-site warehouses | Moderate; each phase is a smaller, contained change |
Big Bang | The whole warehouse switches at one point, usually over a weekend. | Small single-site operations with low SKU counts and predictable volumes | Highest; a failure affects everything at once |
Zero downtime means parallel running or a phased rollout. With parallel running, the old system stays the system of record until the WMS produces identical results. With a phased rollout, a multi-warehouse business can prove the system at one site before moving to the next. Big Bang can work for small operations, but the rollback plan then becomes the only safety net.
Step 5: Test at full volume, train by role, write the rollback plan
Full-volume testing
Simulate a peak day at 110 to 120% of normal daily volume, end to end: receive POs from the ERP, process goods-in and putaway, release and pick orders, pack, label and confirm despatch, then reconcile stock between WMS and ERP. Run it with warehouse staff, not just the project team. Log every issue with an owner and a deadline. Do not go live with open critical or high-severity issues.
Training
The gap between a technically successful implementation and one that works on the floor is almost always a training gap.
Start early. Introduce the system and the reasons for change 6 to 8 weeks before go-live.
Train by role. A goods-in operative does not need despatch workflows.
Use the real system. Test environment, real product data, real scanners, real label printers.
Appoint floor champions. One or two experienced operators per shift with deeper training who act as first-line support.
Staff the first two weeks. Extra support on the floor for at least the first ten working days.
Rollback plan
A rollback plan removes the fear that a failed go-live leaves you stranded. It should cover:
Trigger criteria. Defined in advance: for example, WMS-ERP integration down for more than 30 minutes, stock divergence above an agreed threshold, pick accuracy below an agreed level.
Decision authority. One named person, usually the Operations Director or sponsor. Not a committee.
Rollback steps. If running parallel, stop using the WMS and continue on the legacy system. After a Big Bang, you need a tested data restore.
Communication. Pre-drafted messages for warehouse staff, customer services, carriers and suppliers.
Time limit. Rollback stays a realistic option for the first 48 to 72 hours. After that, commit to fixing forward.
Test the rollback during the full-volume test. Simulate a failure, execute the rollback and confirm you can resume on the old system within the target time.
What this looks like in practice: CMS Distribution
CMS Distribution, a UK technology distributor with a 140,000 sq ft warehouse, went live on Infios (then K.Motion) Warehouse Edge, implemented by Balloon One and integrated with SAP S/4HANA, at the start of Q4, its peak trading period, while consolidating two warehouses into one new facility.
The project delivered broadly on time. Carrier integration removed manual label printing, stock visibility moved from a once-a-day spreadsheet to live data, and the business recorded a 10 to 15% efficiency improvement and a 10 to 15% lower cost per transaction. Jon at CMS Distribution: “I’m not sure we would have so successfully got through that peak time without an automated system.” Full detail in the CMS Distribution case study.
Other Balloon One UK implementations have recorded a 38% picking improvement (Pets Corner), 99.6% OTIF (Rapid Electronics), a 40% reduction in picking errors (Centrado) and a 24% year-on-year inventory reduction worth £2.6 million (Virgin Wines).
Zero-downtime go-live checklist
Every item needs a named owner and a completion date.
Data and integration
SKU master data cleansed, deduplicated and loaded
Location master finalised and validated against physical warehouse
Lot codes and expiry dates migrated and FEFO/FIFO rules tested
Opening stock count completed and reconciled to ERP
ERP integration tested at full daily volume (sales orders, purchase orders, stock sync, despatch confirmations)
Integration error handling and retry queue confirmed working
Two successful trial migrations completed and signed off
Cutover and rollback
Cutover strategy agreed (parallel, phased or Big Bang) and documented
Cutover schedule published with hour-by-hour task list
Rollback trigger criteria defined and agreed by project sponsor
Rollback procedure tested during full-volume test
Decision authority for rollback named
People
All warehouse staff trained on role-specific workflows
Floor champions identified and given advanced training
Go-live support rota in place for first ten working days
Communication sent to customer services, carriers and key customers
Escalation contacts (WMS provider, IT, operations) documented and distributed
Post go-live
Daily reconciliation process defined (WMS stock vs ERP stock)
Issue log and triage process in place
First-week review meeting scheduled
KPIs to track after go-live
KPI | Target or baseline | Review cadence |
|---|---|---|
Pick accuracy | 99.5%+ within the first month | Weekly, then monthly |
Stock accuracy | WMS count vs cycle count | Weekly |
Order cycle time | Order release to despatch confirmation | Weekly |
Lines picked per hour | Against pre-WMS baseline | Weekly |
Mis-picks and returns | Root cause: system, training or data | Weekly |
Share the results with the warehouse team as well as the board. When people see the numbers improving, adoption accelerates.
Next step
If you are planning an implementation and need a business case that gives the board confidence in a controlled, low-risk cutover, explore Balloon One’s warehouse management solutions or get in touch to talk through your timeline.
Sources
Gartner WMS success rate (24%), via Alpine Supply Chain Solutions
Gartner survey of 306 logistics leaders (76% of transformations miss targets), via Supply Chain Xchange
Zebra Technologies WMS utilisation figures, via Contivos
Balloon One, Out of Date: The Legacy UK Supply Chain Driving Avoidable Food Waste, research with 500+ UK supply chain managers
Frequently Asked Questions (FAQ's).
Single-site cloud deployments typically take 8 to 16 weeks. Multi-site, bespoke-ERP or heavily customised projects take 4 to 12 months. Data readiness and integration complexity drive the difference more than the software does.
Yes. Use parallel running or a phased rollout so the warehouse always has a working system, rehearse the data migration at least twice, test ERP integration at full daily volume and hold a tested rollback plan with named trigger criteria for the first 48 to 72 hours.
Parallel running has the lowest disruption risk but doubles the workload for one to four weeks. A phased rollout suits mid-sized and multi-site warehouses because each phase is a smaller change with its own rollback point. Big Bang is only advisable for small, single-site operations.
A clean SKU master, a validated location master, a reconciled opening stock count, and for regulated goods, complete lot codes and expiry dates in a consistent DD/MM/YYYY format. See what data you need before implementing a WMS.
Sage 200 typically uses middleware or a direct database connector. Business Central uses APIs. In both cases the integration must move sales orders, purchase orders, stock adjustments and despatch confirmations, and must be tested at real daily volume with field mapping for units of measure, locations and batch numbers.
Trigger criteria defined in advance, a single named decision-maker, documented steps to revert to the old system, pre-drafted communications for staff, carriers and customers, and a time limit (usually 48 to 72 hours) after which you commit to fixing forward.
January to March or May to July. Avoid October to December peak unless unavoidable, and allow at least 12 weeks of implementation before the target date.
Cost depends on licence model, integration scope, hardware and internal time. Balloon One’s guide to how much a WMS costs and the WMS business case resource cover the components.





