There is a particular kind of SAP programme that passes every technical test, goes live on the planned date, and is quietly judged a failure eighteen months later. The system works. People have simply built their own ways around it.
This outcome is predictable, and it is usually visible long before go-live to anyone willing to look for it.
Adoption is a design constraint
The most reliable indicator is who was in the room during design. When process design is done to a business function rather than with it, the resulting configuration reflects what the project team understood rather than what the work actually requires. The gap surfaces at go-live, when there is no budget left to close it.
Involving the people who do the work has a second effect that is easy to underestimate. They become the people their colleagues ask for help, which is worth considerably more than any volume of formal training material.
Governance that decides things
- A steering group with the authority to settle scope disputes, meeting often enough to matter.
- A single, visible decision log, so that settled questions stay settled.
- Explicit ownership for each process area, held by the business rather than by the programme.
- A standing route for raising concerns that does not depend on someone being brave in a large meeting.
Testing is a business activity
User acceptance testing delegated entirely to a project team is not user acceptance testing. It confirms that the system does what the project believed it should, which is exactly the assumption under examination. Testing carried out by the people who will live with the result surfaces a different and more useful category of problem.
Plan for the weeks after go-live
The period immediately following cutover is where adoption is won or lost, and it is routinely the least resourced part of the plan. Teams are working in an unfamiliar system while still delivering their day job, and the speed at which their questions get answered sets the tone for everything that follows.
Hypercare deserves to be planned with the same seriousness as the cutover itself: named people, clear routes to reach them, and an explicit trigger for handing support back to normal operations once the volume of questions has settled.
Leave a Reply