WMS Implementation Checklist: 35 Steps From Slotting Data to Go-Live Week
A sequenced, owner-assigned plan for funded WMS projects: master data, slotting, label checks, ERP/TMS testing, the inventory freeze, and the go-live metrics that tell you when to step in.
Name Every Owner and Write the Go-Live Gate First
The contract is signed and the budget is approved. Most of the risk left in a WMS project comes from doing things in the wrong order. Teams configure pick rules before item dimensions are right, test integrations with ten clean orders instead of ten thousand messy ones, and pick a go-live date in the week finance closes the quarter. This WMS implementation checklist puts 35 steps in order and gives each one a single owner. It assumes a project of roughly 12 to 24 weeks for one site. Multi-site rollouts repeat steps 6 through 35 for each building.
1. Give each workstream one accountable owner (Owner: executive sponsor). You need named people for master data, integration, devices and labels, floor operations, finance and training. A workstream with two owners has no owner.
2. Set the go-live date around your peaks and your fiscal close (Owner: project manager with finance). Peaks depend on your market: Black Friday, Singles' Day, Diwali, Ramadan and Eid, Lunar New Year, back-to-school. Fiscal year-ends differ too. December is common, but March closes are typical in India and Japan. Leave at least four weeks between go-live and any of these.
3. Write go/no-go criteria now, while nobody is under pressure (Owner: project manager). Examples: every priority-one defect closed, UAT signed by each shift lead, peak test passed, inventory reconciled within the agreed tolerance.
4. Freeze scope in writing (Owner: project manager with vendor or integrator). List the features deferred to phase two, such as labor management, yard management or voice picking, so they don't resurface in week 10.
5. Hold back a contingency and pre-buy hypercare hours (Owner: finance controller). A 10 to 20 percent contingency on services is a common rule of thumb. Implementation services frequently equal or exceed the first-year subscription. Typical 2026 cloud WMS subscriptions for a mid-market site run from low thousands to well above ten thousand dollars per month, depending on users, modules and contract. Confirm with vendors, because regional pricing varies.
| Phase | Typical timing before go-live | Lead owner |
|---|---|---|
| Master data and location build | 16 to 10 weeks | Inventory control lead |
| Slotting analysis | 12 to 8 weeks | Operations engineer |
| Labels and devices | 10 to 4 weeks | IT devices lead |
| Integration and peak testing | 10 to 2 weeks | Integration lead |
| Cutover and inventory freeze | Final 72 hours | Project manager |
| Hypercare | Go-live plus 2 to 4 weeks | Site operations manager |
Item and Location Master Data: Get It Right Before Configuration
Every WMS rule depends on master data: putaway logic, replenishment triggers, cartonization and pick paths. Feed it bad data and the system will make bad decisions quickly and consistently.
6. Extract and profile the item master (Owner: master data owner). Flag missing dimensions, zero weights, duplicate barcodes and inactive SKUs still holding stock. Archive anything with no movement in 18 to 24 months instead of migrating it.
7. Measure the top movers physically (Owner: inventory control lead). Supplier spec sheets are often wrong. Use a dimensioning device or a careful manual team to measure every SKU in your top 20 percent by pick lines, at each and case levels.
8. Define the unit-of-measure hierarchy and a barcode for each level (Owner: master data owner). Each, inner, case and pallet all need quantities and identifiers. GS1 GTINs work in most markets. Internal codes create problems later with retail customers.
9. Build the location master with one naming convention (Owner: inventory control lead). Use zone-aisle-bay-level-position with fixed-width segments, and add check digits so pickers can't confirm the wrong slot by reading a neighboring label.
10. Record location attributes (Owner: operations engineer). Capture cube capacity, weight limit, temperature zone, pick or reserve type, and hazardous-goods segregation class. Storage segregation rules vary by jurisdiction and are set by regional and national codes, so check yours with your safety advisor.
11. Set lot, serial and expiry tracking per item (Owner: quality lead). Pharma and medical device distributors may fall under serialization rules such as the EU Falsified Medicines Directive or the US DSCSA. Food businesses usually need FEFO allocation. These settings are hard to change after go-live.
Slot From Order-Line Velocity, Not From Today's Layout
A new WMS is the cheapest chance you'll get to fix a layout that grew without a plan. Don't re-slot the whole building, though. Fix the high-velocity zones and leave the long tail alone.
12. Run an ABC analysis on 12 months of pick lines, not revenue (Owner: operations engineer). A low-value consumable picked 400 times a day belongs in the golden zone ahead of an expensive item picked twice a week.
13. Find affinity and seasonal items (Owner: operations engineer). Items that often ship together should sit close to each other. Seasonal SKUs need a planned re-slot date, not a permanent forward location.
14. Apply ergonomic and weight rules (Owner: health and safety lead). Heavy, fast items go at waist height. Manual-handling limits come from local regulation and differ by country, so encode your site's limits as slotting constraints.
15. Re-slot only the top movers before cutover (Owner: site operations manager). Hypothetical example: say a 30,000-SKU distributor moves its top 1,500 SKUs over two weekends before go-live. The rest moves through system-directed re-slotting over the following quarter. Changing layout and software together makes it impossible to tell which one is causing problems.
Validate Labels, Printers and Scanners on the Real Floor
Label and scanning failures are a common reason a go-live stalls on day one. They're also among the easiest to prevent.
16. List every label format (Owner: IT devices lead). Include location labels, license plates, GS1-128 SSCC pallet labels, carrier labels for each carrier and service, and customer-specific compliance labels. Large retailers can charge chargebacks or fines for non-compliant labels.
17. Print and scan every format on every printer model (Owner: IT devices lead). Use real stock and real ribbons, and grade the barcodes with a verifier. A printer that's fine for location labels may produce marginal 2D codes.
18. Run a wireless survey of the whole building (Owner: network lead). Freezers, high-bay racking, mezzanines and steel-heavy dock areas tend to have dead spots. Ask whether the WMS mobile client works offline or freezes when it loses signal.
19. Enroll devices in mobile device management and keep about 10 percent spares (Owner: IT devices lead). Typical 2026 prices run roughly $1,000 to $3,000 per rugged handheld and $1,000 to $3,500 per industrial label printer, depending on spec and region. Confirm with suppliers. Plan battery swaps for every shift.
20. Test scanning in real conditions (Owner: shift supervisors). That means gloves, cold-store condensation, scan distances to top-level racking, and worn labels. Bench tests under office lighting don't count.
Integration Testing and Peak Testing With Production-Shaped Data
WMS integration testing fails when teams test the happy path with clean data. Your ERP and TMS will send partial shipments, cancelled lines, address changes and duplicate messages. Test with those.
21. Map every interface (Owner: integration lead). Typical flows are item master, purchase orders and ASNs, sales orders, shipment confirmations, inventory adjustments, returns, and carrier rating and labeling through the TMS or parcel API. Record the direction, trigger, frequency and system of record for each.
22. Unit-test each message, including failures (Owner: integration lead). Malformed payloads, timeouts and duplicates should land in a visible error queue with alerts, not fail silently.
23. Run end-to-end scenarios across ERP, WMS and TMS (Owner: business process lead). Cover order to ship, PO to putaway, returns, transfers and cycle-count adjustments. Where a shipment triggers a statutory document, such as India's e-way bill, Mexico's CFDI waybill complement, or e-invoicing mandates in the EU and Gulf states, include that step and confirm the details with your tax advisor.
24. Prove inventory reconciliation (Owner: finance controller). Compare WMS and ERP on-hand by SKU and location after a full test cycle. Any unexplained variance blocks go-live.
25. Load-test at 1.5 to 2 times your peak day (Owner: integration lead with vendor). Time wave release, allocation, label printing and shipment confirmation. If you're also running a cloud WMS comparison for other sites, ask each vendor how load tests work in a multi-tenant environment. Some limit test volumes in shared sandboxes.
26. Measure carrier and TMS call times under load (Owner: integration lead). A label call that takes four seconds is fine for one order and unworkable when 40 packers are waiting.
27. Have floor staff run UAT and sign it (Owner: shift supervisors). Super-users from each shift run their real tasks. Sign-off from the people who do the work counts for more than sign-off from the project team.
Build the WMS Cutover Plan Around a Physical Inventory Freeze
A project that went well can still fail at cutover. Write the plan hour by hour and rehearse it at least once.
28. Write an hour-by-hour cutover plan with a named rollback point (Owner: project manager). Each task gets an owner, a start time and a check. Agree in advance on the last hour at which you can still revert to the old system, and who makes that call.
29. Schedule and communicate the freeze (Owner: customer service lead). Stop receiving and shipping for roughly 24 to 72 hours, depending on how many SKUs you carry. Tell customers, suppliers and carriers two to three weeks ahead and reschedule inbound appointments.
30. Count the inventory you'll load (Owner: inventory control lead with finance controller). Run a full wall-to-wall count, or count only the zones that fall short of your cycle-count accuracy target. Load the counts into the WMS, reconcile against the ERP, and have finance sign the variance write-off. Your auditors may want to observe the count.
31. Decide which open orders migrate (Owner: business process lead). Finish and ship already-picked orders in the legacy process before the freeze. Unreleased orders move to the new system. Partially picked orders cause the most trouble, so clear them first.
32. Staff super-users on every shift and plan overtime legally (Owner: HR and site operations manager). Rules on overtime, rest breaks and night work vary by country and sometimes by state or province. Check them before you schedule a 72-hour cutover.
Go-Live Week Metrics and When to Intervene
Expect productivity to fall in week one. What matters is whether it recovers each day. Watch a small set of numbers with agreed thresholds instead of reacting to individual complaints.
33. Hold a war-room stand-up at every shift change (Owner: site operations manager). Keep it to 15 minutes: the metric board, the top five issues, an owner and a deadline for each.
34. Track these metrics against your pre-go-live baseline (Owner: operations analyst).
| Metric | What it tells you | Suggested trigger for action |
|---|---|---|
| Picks per labor hour vs baseline | Whether the floor is adapting | No day-over-day improvement for two consecutive days |
| Orders shipped on time | Customer impact | Below the service level in your customer contracts |
| Integration error queue depth | Silent data drift between systems | Any message older than one hour |
| Inventory accuracy on cycle counts | Whether transactions are recorded correctly | Below your agreed tolerance in any zone |
| Short picks and exceptions per shift | Slotting or replenishment faults | Rising trend after day three |
| Dock-to-stock time | Receiving and putaway health | Inbound backlog older than one shift |
35. Exit hypercare against written criteria (Owner: executive sponsor). Typical criteria: productivity within an agreed band of baseline, no priority-one defects open, an empty error queue for five straight days, and reconciliation clean through one weekly close. Then apply this rule: if on-time shipping and the error queue are both healthy, keep going and work through the productivity dip. If either is failing on day four, stop new scope work and put the vendor's team on that problem until it's fixed.
Quick answers
How long does a typical WMS implementation take?
A single-site cloud WMS project usually takes about 12 to 24 weeks from kickoff to go-live. The main factors are integration complexity, how clean your data is, and automation such as conveyors or sorters. Projects with heavy material-handling automation or several ERP instances often run longer. Ask your vendor or integrator for a phase-by-phase plan, not just an end date.
Do we really need to stop shipping for a physical inventory count at cutover?
Usually yes, at least for the zones where you can't trust current accuracy. A WMS that starts with bad balances sends pickers to empty locations from the first shift. Sites with consistently high cycle-count accuracy can sometimes count only the weaker zones and shorten the freeze to about 24 hours.
What should WMS integration testing include beyond basic order flows?
It should cover error handling, duplicate and out-of-sequence messages, cancellations, partial shipments, returns and inventory adjustments. It should also include any statutory documents a shipment triggers in your countries, such as e-way bills or e-invoices. Finish with a full reconciliation of WMS and ERP on-hand by SKU and location.
How much productivity loss should we expect in go-live week?
Some drop is normal while staff learn new screens and the system directs them to new locations. The size depends on training, slotting changes and device readiness. Don't fix a percentage in advance. Track whether productivity improves day over day, and escalate if it stalls for two consecutive days.
Should the WMS vendor or a third-party integrator run the implementation?
Vendor-led teams usually know the product best. Independent integrators are often stronger on ERP and TMS integration and can bring local language and regulatory knowledge in your region. Either way, check that the named consultants have delivered projects of similar scope, and put hypercare hours and defect response times in the statement of work.
Written by the Fleetmesh Tech editorial team. Prices, features and rules change often - confirm details with the vendor or a licensed professional before you decide.