Somewhere in your network it is always the middle of a shift.
Global testing strategy and follow-the-sun cutover command across multi-region deployments. The recurring lesson: in a network that never pauses, the quality of the command model decides the cutover, and the plan is only the thing you argue with. Walk the five decisions, find your seat, then run the two-minute check.
Five calls that decide a logistics program
Made before cutover, tested during it, and visible in service levels within days. Pick a decision.
A multi-region cutover runs across time zones, and control passes between teams several times. Most plans describe tasks and not authority.
Define who holds go and no-go authority at every hour of the window, and write the handoff protocol before the window opens.
Cutover failures in a follow-the-sun model are almost never technical. They are decisions made twice, or not at all, because control passed without context. The handoff protocol is the deliverable, not the task list, and it should specify what gets said, by whom, and what evidence travels with it.
Who has go and no-go authority at 03:00 in each region, by name?
Each region tests what it knows, which produces good local coverage and no evidence about the flows that cross regions.
Own one testing strategy centrally with regional execution, and make cross-region flows a named scenario set rather than an assumption.
The defects that hurt are the ones between regions: a shipment created in one, consumed in another, with different master data and different local practice. Nobody owns those scenarios by default because they belong to no single region, which is exactly why they survive testing and appear in production.
Which test scenarios deliberately cross a regional boundary, and who owns them?
The ERP sits inside a network of warehouse, transport and carrier systems, plus customs and partner interfaces that were built at different times by different people.
Inventory every interface with an owner per line, and name a client-side owner for the seams rather than an owner on each end.
This is where the program lives. Sequencing and master data mismatches across those seams produce shipments that exist in one system and not another, and the operational answer is a manual workaround within hours. Once that exists your data stops describing the network.
Can we produce the full interface inventory with an owner per line today?
Locations, items, units of measure and partner records have accumulated regional variation, and each version is correct locally.
Baseline master data quality before design locks and scope remediation as a program with an owner.
Cross-region flows fail on exactly this. A unit of measure that means one thing in one region and another elsewhere produces quantities that are individually plausible and collectively wrong. Discovered at integration testing, there is no schedule left to absorb the cleanup.
What is our measured master data consistency across regions today?
The go or no-go conversation arrives and nobody can say what not going live costs, because nobody priced running the legacy estate another quarter.
Price the fallback before accepting a date, and write the contingency plan while there is still time to fund it.
Without a priced alternative you accept a date you do not believe, because there is nothing else to point at. With one, the go decision becomes a comparison, and the conversation with your integrator changes tone immediately. The double-run cost is also the line most often missing from the business case.
What does running the legacy estate another quarter cost, and who decides?
What those five mean for the chair you sit in
Logistics programs get judged on service levels within days. Each seat's exposure, the early sign, and the question worth asking this quarter.
Workarounds appear within hours
When a shipment exists in one system and not another, the network fixes it manually because freight has to move. That is correct operationally and corrosive to your data. The other exposure is the cutover itself: authority passing between regions without a protocol produces decisions made twice or not at all.
The cutover plan lists tasks and does not name who holds authority hour by hour.
Who holds go and no-go authority at each hour of the window?
Cost per unit depends on data crossing seams
Landed cost, cost to serve and margin by lane all read from data that crosses systems the program does not own. Master data variation across regions produces numbers that look plausible and are not. And if nobody has priced running legacy another quarter, you have no position in a schedule argument.
No priced fallback exists for a delayed go-live.
What is the priced cost of not going live on the current date?
Shift and cross-border workforce complexity
Warehouse and driver populations run rotating shifts, premium patterns and in many networks separate agreements per site. Generic templates assume one schedule and one manager. Add the platform clock: on-premises Workforce Central reaches end of life in March 2027, and moving off it is a reimplementation rather than an upgrade.
Hypercare exit is measured in weeks rather than in completed payroll cycles across a full shift rotation.
How many parallel payroll cycles cover a full rotation at our most complex site?
A cutover in a network that never pauses
Decision one, drawn out. Control passes several times inside a single window. What travels with it decides whether the next region starts from evidence or from assumption.
Cutover failures here are rarely technical. They are decisions made twice, or not at all, because control passed without context.
Four items already on the calendar
None of these are logistics-specific, and all four land on a network program. Verify each against your own estate before the next steering meeting.
Engineering stopped at the end of 2025. Moving off it is a reimplementation rather than an upgrade, and any shift-based workforce is in scope.
Extended maintenance runs to 2030 for a fee. If the target platform is a lift and shift of ECC, it arrives with a published expiry date attached.
Oracle support runs past 2036. When an integrator sells urgency on that basis, the pressure is customization debt and scarce skills, not vendor abandonment. Knowing the difference is a negotiating position.
Not a deadline, a treadmill. Two mandatory regression cycles annually, permanently, and the item most reliably missing from a post-go-live staffing plan.
Five questions worth more than a readiness assessment
Answerable from memory, scored on this page, nothing captured and nothing emailed.
These five are the start of the instrument. A full review also covers the carrier and customs interface set, cross-region master data governance, command staffing and hypercare exit criteria. Or skip the tooling and book the program review.
Cutting over a network that never stops?
Pre-SOW, mid-build, or stabilizing after a rough regional go-live. I sell no software and staff no builds, so these questions get asked out loud. Tell me where the program is and I will tell you what I see.