The 2 AM Phone Call
2:17 AM, September 12, 2023 — WMS go-live night, my first hard lesson in WMS go-live training.
My phone rang. It was Lao Zhao, the night-shift lead, stuttering so bad I could barely make out the words: "Feng, the system… it's frozen. Outbound orders won't release. The drivers are cussing us out at the dock doors."
Our new WMS had gone live at midnight that day. The daytime cutover looked smooth. Then the night shift took over and everything fell apart. The system wasn't broken — the people didn't know how to use it.
I dug into it afterward. Of the 30-plus people on that night shift, only about half had attended the pre-go-live training. The rest had "heard about it" from coworkers. Over 60 mis-shipped parcels in one night.
From that day on, my hard rule: the two weeks before a WMS go-live, training is the number-one priority. Even the shift schedule bends around it.
Three Go-Lives, Three Lessons in WMS Go-Live Training
Full disclosure: I've run a 200-person operation through three WMS implementations. First: disaster. Second: half a disaster. Third: finally right. Every time, the difference wasn't the software. It was the training.
Round one, 2019, a 30,000-square-meter e-commerce DC. The vendor sent two consultants who lectured off PowerPoint for three days, then handed out an 80-page user manual. On go-live day, pickers mistook "wave release" for "print shipping labels" and cranked out 4,000 wrong labels before lunch. Two full days of rework. Six figures down the drain, easy.
Round two, 2021, an apparel warehouse. We ran "all-hands training" — everyone from pickers to supervisors in one room, all sitting through the exact same course. Pickers endured a full morning on inventory transfer strategy. After go-live, supervisors could read dashboards; the floor crew still couldn't operate the handhelds. Same mistakes, new system.
Round three, 2023 — the one from the opening story. We crashed at first, then rebuilt the whole training plan in two weeks and re-launched. That relaunch became the model rollout across the company. We've used the same playbook ever since.
The Counter-Intuitive Truth: Training Isn't About Teaching the System — It's About Changing Habits
A lot of folks think WMS go-live training means teaching people which button to tap. Wrong.
Here's the honest truth: a picker with eight years on the floor learns the buttons in two days. The real problem: he used to memorize locations in his head, and now he has to scan; he used to shout across the aisle, and now he has to file an exception in the system. That's a habit problem, not a skills problem.
So our goal was never "everyone can operate it" — it's three levels, in order:
- Dare to use it — not afraid of tapping the wrong thing, and knows how to back out when they do
- Can use it — completes their own job's full loop independently
- Uses it smoothly — when something abnormal happens, knows who to call and which path to take, instead of improvising
Number three matters most: in week one, 80% of chaos comes from mishandled exceptions, not the normal flow.
And there's a trade-off I need to lay out plainly: training time versus production time. Pull people into training and output drops — we measured roughly 15% of labor hours lost to off-floor training in a 200-person building. But the cost of rework, mis-ships, and customer complaints after a bad launch runs 3 to 5 times the training cost. Management has to do that math first, or the training schedule never survives.
The 14-Day WMS Go-Live Training Plan: This Is How We Do It Now
The table below is the version we locked in after round three. Role-based, phased, 14 days total. Take it, tweak the numbers, use it.
Phase 1: Train the Seeds First (Days 1–3)
Start with your "seed" crew: 2–3 sharp, respected veterans from each shift, plus every shift lead and supervisor. About 30 people total.
| Day | Content | Format | Pass Criteria |
|---|---|---|---|
| Day 1 | End-to-end flow walkthrough: how a document travels from receiving to shipping | Consultant lecture + tabletop simulation | Can sketch their own step's upstream and downstream |
| Day 2 | Full hands-on for their own role: run the normal flow three times | Live practice in the test environment | Completes the loop independently, no manual |
| Day 3 | Exception handling: count variances, damage, stuck orders — what now? | Scenario drills; the consultant deliberately sets traps | 8 out of 10 exception scenarios answered correctly |
These 30 become the instructors. On day 3 afternoon we re-screen: anyone shaky or unable to explain gets swapped out. A half-baked instructor sinks everything downstream.
Phase 2: Rolling Crew Training (Days 4–10)
Split the 200-person crew into 4 batches of 50, each training 2 days, rolling. Production never stops: every shift keeps experienced hands on the floor.
| Block | Content | Format |
|---|---|---|
| Batch day 1, AM | Role-specific operations only — nothing extra | Instructor-led, one handheld per person in test env |
| Batch day 1, PM | Exception reporting: when to stop, who to call, how to log it in the system | Real case reviews — we use our own past screw-ups as material |
| Batch day 2, AM | Full simulation: run a peak-season order volume end to end | Stress test, order volume loaded at 1.5× peak |
| Batch day 2, PM | Test + Q&A | Hands-on exam + open floor questions |
Pass criteria by role:
| Role | Must-pass items | Bar |
|---|---|---|
| Picker / checker | Normal pick flow + exception reporting | ≤2 errors across 20 hands-on steps |
| Receiver | Receiving + putaway + variance handling | Handles 3 common variance types independently |
| Shift lead | Full flow + dashboards + crew scheduling | 80+ on both written and practical |
| Supervisor | Reports + exception approvals + rollback plan | Can state the rollback trigger conditions cold |
Two failures means no solo shifts in week one — they pair with a veteran. Harsh, but on go-live day you'll be glad.
Phase 3: Dress Rehearsal (Days 11–14)
| Day | Content |
|---|---|
| Days 11–12 | Full integration run: two days of real historical orders (sanitized) in the test environment; shifts take turns playing "find the flaw" on each other |
| Day 13 | Cutover rehearsal: simulate the midnight switch — old system down, new system up, data doesn't reconcile, now what? Everyone walks it |
| Day 14 | Review + rally: walk the full issue list from the past 13 days, assign every item an owner; team dinner that night, morale maxed before launch |
Day 13 is "the dress rehearsal." In 2023 it caught mismatched printer drivers between test and production — gibberish labels. Imagine that on go-live morning.
Three Unfancy Practices We Never Dropped
First, veterans teach rookies — no giant lecture halls. Two hundred people in one room means half are on their phones. Our rule: one instructor, max 8 people, right on the floor, handheld in hand, teaching between picks. Learn it, use it immediately.
Second, morale is part of training. Veterans especially fear the new system makes them expendable. So in the week before launch, management talks to every veteran one-on-one: "The system works for you, not instead of you. Master it and you're the most valuable person in this building." Plus a month-one "System Star" award — five winners a week, cash bonus, name on the wall. Small money, big face.
Third, keep a rollback plan in your pocket — but don't tell everyone. If core flow success drops below 90% within 4 hours of cutover, or any link jams completely, we flip back. Only supervisors and above know the triggers; the floor crew just charges forward. Why? If people know there's a way back, they don't learn with urgency. But management must hold that safety line.
One Reminder for Fellow Operators
A WMS selection costs millions, implementation takes half a year — and where it dies is the cheapest, least respected part: these two weeks of training.
The system is dead code. People are alive. A system truly goes live the moment the last employee dares to tap "confirm" on their own. Until then, every project timeline is just paper.
Every new rollout here still follows this 14-day plan. Adjust the numbers — but "seeds first, rolling crew training, dress rehearsal," carve those on the wall.
Don't let the system fail on "nobody knows how to use it." I paid for that sentence with three warehouses' worth of lessons.
By Chen Feng, Warehouse Operations Manager
