Here is the short version: a wms go live first week war room is not a room where people hold meetings. It is where problems get fixed while orders keep moving. I have run more than 30 go-lives, and every first week that turned into chaos had the same root cause. Nobody had written down, before go-live night, who decides what, who fixes what first, and when we roll back.
The week after cutover is the most fragile week a warehouse will ever have. Productivity typically drops 30 to 50 percent in the first week, and that is normal. It recovers within two to three weeks if the support structure holds. The war room exists to make sure it holds.
Last March I was on site at a 3PL warehouse in Ontario, California. Sixty thousand square feet, about 8,000 orders a day. At 7 a.m. on go-live morning, the first picker scanned the first order and got an error: the location code on the shelf did not match the location in the system. Forty people stood around waiting. Two consultants were in the server room. The warehouse manager was on the phone with a client. The project manager was posting in a group chat asking who could take a look. Three hours later we found it. During the pre-cutover count, someone had changed more than 200 locations by hand, and nobody synced them into the system. Two hundred locations, three hours, one whole day of shipping plans wrecked. I have seen this exact failure three times. Each time, the war room had no one with authority to decide on the spot, and no list that said what gets fixed first.
Who Sits in the WMS Go Live War Room, and Why Each Seat Matters
Five roles, minimum. Lose one and the first week will find the gap for you.
First, the floor commander, usually the warehouse manager or operations director. One power: when something breaks on the floor, their word is final. Stop the line, change the process, no escalation needed. Second, the system lead, usually the consultant on site, who can change configuration and parameters directly. Third, the integration and data owner, who watches the ERP order feed, the inventory sync, and the carrier label printing. Half the systems that die in week one die at an interface. Fourth, super users, at least two per shift, the floor staff who know the new process best. They answer questions on the spot so the consultants do not drown in "how do I do this" tickets. Fifth, the duty manager for nights and weekends. There must always be someone who can make a call at 2 a.m. "Wait until tomorrow morning" is not an answer in go-live week.
Sit together, physically. A chat group is not a war room. Two hundred messages scroll past and the one that matters gets missed. Put the room inside the warehouse, two minutes' walk from the pick floor. Whiteboard, printer, spare scanners, a spare laptop. The door stays open all week.
One role people forget: the client communicator. In a 3PL warehouse, orders will be late in week one and clients will call. If the warehouse manager is firefighting and answering client calls at the same time, both jobs fail. Assign one person to handle client calls and send an hourly progress update. That frees everyone else to fix the actual problems.
How the War Room Triages Issues
Four severity levels, written on the whiteboard where everyone can see them. Everyone memorizes them before day one. In a wms go live first week war room, this triage table is the only thing standing between an orderly fix and a shouting match over whose ticket goes first.
Severity one means the line is stopped: system down, orders not flowing, the whole pick line idle. Someone starts fixing it within one hour. If it cannot be fixed, the rollback plan starts. Severity two means a batch failure: a whole wave of orders wrong, a whole zone's inventory off, an interface failing in bulk. A workaround must exist within four hours, manual if necessary, so freight keeps moving. Severity three means a single-point problem: one order error, one bad location, one operator stuck. Super users handle it on the floor and log it the same day. Severity four means an enhancement: a screen could be nicer, a report is missing a column, a flow could be faster. In week one, severity four gets logged and nothing else. It waits until week two.
The triage table exists to stop the loudest voice from jumping the queue. I watched a warehouse where the finance director called on day two about a report format, and the consultant dropped a broken interface fix to rework the report. That afternoon the interface failed again and more than 300 orders missed the carrier cutoff. A severity table is what lets you say no to that call.
Two stand-ups a day, fifteen minutes each, no exceptions. At 7:30 a.m.: how many severity one and two issues are still open, what is today's ship target, which zone needs extra eyes. Before close: how many issues were resolved, what remains, who is on duty tomorrow. Three things only: numbers, owners, deadlines. Anyone telling a backstory gets cut off.
The Rollback Plan Goes on Paper, on the Wall
Many projects have a rollback plan that says one sentence: "If it really fails, we switch back to the old system." That is not a plan. A real rollback plan has four things written down.
First, triggers, in numbers, not adjectives. For example: system down for more than two hours; order backlog above 30 percent of the daily plan; inventory accuracy below 95 percent. Hit any one and the floor commander can trigger rollback immediately, no meeting required. Second, the steps: where the old system's data comes from, how paper pick lists get reissued, how already-shipped orders get reconciled. Write them out and rehearse them on cutover night. Third, the rollback owner and phone number, on the whiteboard. Fourth, the recovery plan: when we cut back to the new system, who verifies it, and how.
In September 2024 I was at a go-live in Texas where the ERP interface died for four hours on cutover night and the backlog hit 1,500 orders. The floor commander followed the plan: paper pick lists, manual verification, back-entry after the system recovered. By noon the next day the backlog was cleared. Clients got one delay notice and not a single lost order. A rollback plan is not valuable because you use it. It is valuable because when something breaks, nobody panics. A plan that has never been rehearsed is the same as no plan at all.
The Daily Checklist for the WMS Go Live First Week
Before anyone leaves each day in week one, the war room runs through a checklist. Every item gets a tick. Nobody goes home with a blank box.
Inventory reconciliation comes first. Count a sample of physical stock against the system. If more than five SKUs are off, find out why the same day. Bad inventory in week one poisons everything downstream. Second, order status. How many orders are still open, where exactly are they stuck, and who owns each one. Third, interface health. Run a test order through the ERP feed, the inventory sync, and the label print. All three must be green. Fourth, equipment. How many scanners and handhelds are down, and are there enough spares. Fifth, people. Which shift has the highest error rate, and does it need another round of floor coaching tomorrow.
One more number to watch daily: the productivity recovery curve. Plot shipped orders per hour, every day. A dip of 30 to 50 percent in week one is expected, and it should climb back within two to three weeks. The danger is not the dip. It is a dip that does not climb back. Two flat days in a row means something is wrong: training, process, or configuration. Find it then, not in a month-end review.
If you take one thing from this piece, take this: a wms go live first week war room is judged by how boring the week feels. When the room is prepared, problems get fixed quietly, one by one, and nobody remembers the drama. Boring is the highest praise a go-live can get. I have seen that Ontario failure three times now. I would rather not see it a fourth.
By Lin Xiao, WMS Implementation Consultant







