WMS Go-Live Plan: The Day-Of Sheet

This page covers one day: the day you switch. Project phases, data cleanup and testing are in the implementation checklist by phase. By the time you open this sheet, those should be done.

A go-live goes wrong in small ways, mostly. A printer is on the wrong network. Two people pick from the same pallet. Nobody knows who may override a short receipt. A plan written in advance means these become items on a list, not surprises.

Step 1: choose the window

Pick the date by looking at your own calendar, not the vendor's.

Avoid month-end if you bill from the system. Invoices are hard enough without a new system behind them.
Avoid your peak week and any customer promotion you know about.
Prefer a stretch where the dock is quiet, such as after the last carrier pickup on a Friday, with a weekend to count and load.
Check that the people you need are not on leave. That includes the vendor's contact, your project owner and your best floor lead.

Write the date, the cut-off time and the first live shift on one line and send it to everyone, including customers who will notice a delay. A 3PL should tell clients what happens to orders received during the switch.

Step 2: freeze and count

The cut-off is the moment the old system stops being the record. Everything after it goes into the new one.

Announce the freeze time. After it, no receipts, moves or picks in the old system.
Finish or park work in progress. Orders half picked, pallets on the dock and returns waiting on a bench each need a decision, written down and visible: finish in the old way, or restart in the new system.
Run the final count of the locations you agreed to count. Two people per area, one counting and one recording, works better than one person doing both.
Load the opening stock from the count, not from the last system report. The count is the truth. The report is what the old system believed.
Compare the loaded stock to the count for a sample of locations before anyone picks. If a location is off, fix it now.

A short example, hypothetical: a team finds on Saturday afternoon that eleven locations in the mezzanine were counted but never loaded. Because the loaded stock was checked against the count sheets before opening, they caught it that evening. Without the check, the first pickers on Monday would have found empty locations, and the first reaction would have been to blame the system.

Step 3: staff the floor

Write a roster for the first live shifts. Names, not roles.

Who	Where	What they handle
Project owner	Office, walks the floor each hour	Decisions, escalations, the rollback call
Floor lead, each shift	On the floor	Who does what, breaks, exceptions
Super user, each shift	Beside the busiest station	Questions about screens and scanners
Vendor contact	Reachable by phone or chat	Bugs, settings, data fixes
IT contact	Reachable	Network, printers, devices, logins

A few rules that help:

Super users help, they do not do the work for the picker. If they take over the scanner, nobody learns.
Keep a single list of issues, on a whiteboard or a shared sheet, with a time, a station and a short description. Do not rely on memory.
Decide in advance who may override what. A short receipt, a stock adjustment, a price change: name the person who can approve each one.
Tell people it is fine to slow down. Speed on day one is not the goal.

Have paper ready for the worst case: printed pick lists from the loaded data, blank count sheets and labels. You will probably not use them.

Step 4: run the first orders in a controlled way

Do not release the whole backlog at once. Release a small first wave, ship it, and look at the result.

Start with a few simple orders, then one or two awkward ones, such as a split line or a customer with special labels.
Follow each of the first orders from allocation to carrier handover. Watch what the screens ask and where people pause.
Check the documents that leave the building: labels, packing slips, ASNs, invoices. A wrong label costs more than a slow screen.
Receive the first delivery slowly, with the project owner standing there.

Increase volume only when the previous wave went through without a stop. If waves keep stalling, that is useful information for the rollback decision below.

Step 5: agree the rollback trigger first

Decide this before the switch, with the people who can approve it, and write it down. A trigger is a condition and a time, such as:

"If we cannot ship the first wave by 2pm, we stop and review."
"If stock in more than a set number of sampled locations does not match the shelf, we pause picking."
"If labels cannot be printed to the carrier's standard, we stop outbound."

Then write what stopping actually means. Does the team return to the old system, or to paper, or hold orders until a fix? Going back is only possible if the old system is still intact and the freeze has not been undone, so keep it read-only and untouched until you are well past the trigger.

Most go-lives never use the trigger. Having it calms the room, because people know that a bad hour does not mean a disaster.

Step 6: the end-of-day review

Hold a short meeting at the end of each live day while it is fresh. Fifteen minutes with the project owner, floor leads, super users, the vendor contact and IT.

Go through the issues list and mark each item one of three ways:

Fix tonight: blocks tomorrow's work.
Fix this week: annoying but workable.
Training: the system is right and the person did not know.

Then:

Count a sample of locations and compare to the system. Do the same each morning.
Check what shipped against what was ordered.
Agree tomorrow's roster and the volume you will release.
Send a two-line note to the rest of the business: what worked, what is still being fixed.

If you bill clients from the system, compare the first days' activity to your old records before any invoice goes out. Do not send the first invoice on trust.

After day one

Keep the extra support in place until the issues list is short and the morning count matches the system. Treat that as the exit condition, not a number of days.

When it settles, write down what you would do differently next time. If you are still choosing a system, the RFP template is where to ask vendors how they support go-live and how much of their time you get on the day.

Questions people ask

When is the best time to go live with a WMS?

Pick a quiet stretch of your own calendar, away from month-end billing, peak season and big customer promotions. Many teams start right after the last shipping cut-off of a week so the switch happens while the dock is quiet.

Who should be on the floor during WMS go-live?

Your project owner, the floor lead for each shift, one super user per shift, and a named vendor contact who answers the phone. Everyone else should know who to go to before they ask a colleague.

What is a rollback trigger?

A condition you write down before go-live, such as "if we cannot ship the first wave of orders by 2pm, we return to the old process". It turns a panicked argument into a check against a line you already agreed.

How long should go-live support last?

Keep extra people on the floor until the daily issues list is short and the morning count matches the system. That is a result, not a number of days.