Direct answer
The count you start with decides whether the system is trusted. If the opening stock in a new WMS is wrong, every later discrepancy is blamed on the software, and people drift back to their old habits. A good go-live is mostly preparation: clean master data, labelled locations, an accurate opening count, trained users and a clear plan for the cutover day and the weeks after.
This checklist is a planning aid, not a schedule. We do not put fixed durations on it, because they depend on your size, products and starting data, and we make no hardware certification claims.
Master data
The system is only as good as the product and partner data it starts with. Review it before loading, not after.
- One record per product, with a unique code, description and unit of measure
- Barcodes captured and checked for duplicates, where two products share one code
- Batch, lot, serial or expiry tracking decided per product, before any stock is loaded
- Suppliers, customers and, for a 3PL, the stock owner for each item
- Retired or duplicate products cleaned out, so they do not clutter search and picking
- A named owner who approves changes to master data after go-live
Location and label readiness
Physical readiness is where projects slip. Every storage position should have a unique code that matches the system's layout, printed and fixed in place before go-live. Item and pallet labels should be available for what you receive and store. Test labels by scanning them in real conditions: at the height, distance and lighting your operators will face. Decide what happens to unlabelled stock and who prints missing labels. Scanning mechanics are described on the barcode scanning workflow page. Confirm any devices against your actual workflow. We do not certify hardware.
The opening stock count
The opening count is the single most important task. Plan it as a project, not a favour.
- Freeze or record all stock movements during the count, so quantities do not change underneath it
- Count by location, with two people or a recount for high-value and doubtful locations
- Reconcile the count against your existing records and investigate large differences before loading
- Decide what happens with stock that is found but not in your old records, and the reverse
- Load the counted quantities into the WMS by location, then verify a sample of locations by scanning them
- Record who counted and when, so questions have answers
Do not load guesses
If a quantity cannot be verified, keep it visible as unresolved rather than loading a confident-looking number. An honest unknown is easier to fix than a plausible error.
User roles and training
Define who can do what before training: who receives, who picks, who adjusts stock, who approves and who administers. Restrict adjustments and approvals to the right people from day one. Train people on the real tasks they perform, using real product data and real labels, ideally in a test environment before go-live. Include the exceptions: a barcode that will not scan, a short pick, a wrong delivery. Name a few confident operators as first-line helpers on the floor. Discrepancy handling is worth its own session; see the warehouse inventory discrepancy workflow.
Parallel run or freeze
There are two common cutover approaches, each with a cost. A parallel run keeps the old process and the new system running together for a period, which is safer but doubles the effort and invites double entry. A freeze-and-switch stops movements briefly, loads the opening stock and starts on the new system, which is cleaner but leaves less margin for error. Choose according to your tolerance for disruption and the accuracy of your opening data. Write down the decision, the criteria for stopping the old process and who has authority to call it. Also define what happens if you have to roll back.
Go-live day
Keep the day boring by preparing everything else. Confirm the opening stock is loaded and spot-checked, people and devices are ready and support is present on the floor. Start with a small set of real transactions and watch them end to end: receive, put away, pick, pack, ship. Keep a visible issue log and triage it in real time. Do not fix by editing the database silently; record what changed and why. Decide in advance what counts as a stop-and-fix problem versus a note for later.
The first two weeks
Go-live is the start of adoption, not the end of the project. In the early period, review adjustments and reason codes daily, count high-risk locations soon after movement, and collect operator feedback while it is fresh. Fix recurring exceptions at the process level, such as a label problem or a confusing screen, instead of relying on workarounds. Watch for people returning to paper or spreadsheets, since that is the first sign of trust slipping. Plan a review at the end of the period to decide what to change.
Where LATYNEX fits
Our warehouse management system work is implementation-led. Cutover planning, data preparation and the exact scope are settled per engagement, not assumed. To understand what drives the effort, read about WMS implementation cost, and use the WMS readiness assessment to see where you stand before planning a go-live.
Questions
How long does a WMS go-live take?+
It depends on your size, products, data quality and layout, so we do not give fixed durations. Scope is confirmed per engagement.
Should we do a parallel run?+
It reduces risk but doubles effort. A freeze-and-switch is cleaner but needs a reliable opening count. Decide based on your tolerance for disruption and data quality.
What is the most common cause of a failed go-live?+
An unreliable opening count and untested labels. Everything after inherits those problems.
Do you certify our scanners or hardware?+
No. Devices are confirmed against your workflow, and we make no certification claims.
Who should approve stock adjustments after go-live?+
A small named group, with thresholds set by your operations and finance leads, and never the person making the adjustment.