I’ve been in nine go-live weekends as a director. The one where everything went wrong taught me more than all the smooth ones combined. The budget was blown, the consultants were pointing fingers, and the project manager was crying in the server room. ERP Consulting came in not to fix the software, but to rebuild the decision-making chain under fire.
The 24-Hour Rule That Saved the Project
In the middle of data migration failure, the team wanted to delay the go-live by a month. The cost of delay? $300,000 in lost productivity and a blown quarterly forecast. ERP Consulting proposed a hard rule: make any critical decision within 24 hours or the default is to proceed as planned. This prevented analysis paralysis. We chose to go live with a dirty data set and clean it post-go-live. It was painful but cheaper than delaying.
Three War Room Behaviors That Actually Work
- Use a single “issue tracker” document – not email. Every problem gets a priority (P1 = stop work, P2 = fix within 4 hours, P3 = can wait until Monday). Stick to it.
- Assign a dedicated “naysayer” for every decision. Their job is to argue the worst-case scenario for five minutes. Then the team decides. This kills groupthink.
- Keep a physical whiteboard with “wins” and “lessons learned” updated every hour. It sounds cheesy, but it keeps morale up when the network crashes at midnight.
The One Metric That Tells You Go-Live Is Succeeding
Don’t count how many transactions processed. Count how many are manually corrected within 7 days of go-live. If that number drops week-over-week, you’re out of the danger zone. Ours started at 600 per day, dropped to 200 by day 5, and to 20 by day 30. ERP Consulting called that the “real go-live date.” The next weekend you’re facing a high-stakes cutover, remember: the software is dumb. The people in the room are everything. Give them a process, not a prayer.