Learning path
Running an ERP implementation: ten stations, signature to handover
This is not an explanation of what happens on an implementation — that is written in this site's articles, linked at the foot of the page. This is a path through running one: the document written at each stage, the decision taken there, who signs it, and within what deadline. Finish it and you hold a set of instruments rather than a general understanding.
The path is built around one recurring observation: projects do not stall on technology. They stall because a decision sat unmade for three weeks, because forty small requests were never measured, and because the key user who decides the design is the same person running the business and nobody freed up their time.
- Stations
- 10, in order
- Written for
- A project manager, a consultant, or a sponsor on the customer side
- Assumed knowledge
- The four document cycles
Why the order is not a suggestion
Each station produces a document the next one assumes. The assumptions register in station one is what makes a change request in station three uncontroversial. The success criterion in station two is what gets measured in station ten. Start at testing and you find you have nothing to test against: an expected result that should have been written at design.
Station four — team capacity — is what actually sets your go-live date. It is the one station a project manager cannot solve alone, which is why it is put as a written question to the sponsor in week one rather than as a complaint in month four.
The path, in order
- 1Before the kickoffReading what was signed as a constraint on your project, and a written, approved assumptions register before the first formal meeting.
- 2The charter and governanceAn approved two-page charter, a written decision structure with deadlines, and a decision log started on day one.
- 3Scope and change controlAn approved scope baseline, and a change mechanism that measures every request against time, cost and risk before accepting it.
- 4The plan and the teamA plan built on real capacity in hours rather than on hope, and a written decision on freeing key users before design begins.
- 5Design and the fit-gap decisionRunning design workshops that produce decisions rather than wishes, classifying every gap on a written ladder, and approving a design that was seen rather than read.
- 6Data migrationA migration plan with an owner per file, three load cycles against written reconciliation criteria, and a declared decision on how much history to bring.
- 7Testing and user acceptanceWritten test scenarios with expected results, a daily defect triage, and a numeric exit criterion approved before testing begins.
- 8Training and adoptionA role-based training plan with a measured competence standard, a super-user network chosen against criteria, and adoption indicators read from the system after go-live.
- 9Cutover and go-liveA cutover plan by the hour with owners, a full rehearsal executed and measured, and rollback criteria approved before the night rather than during it.
- 10Hypercare and handoverA numeric hypercare exit criterion, a documented handover to the support team, and measurement of the success criterion written into the charter.
What is being run — four pages
This is a management path rather than a content one. What happens inside each stage is set out where it belongs, and each station links to it once instead of restating it worse:
- ERP team roles — who does what
What each role is and what is expected of it, before you plan anybody’s capacity.
- Data migration — what moves and what stays
The content itself: opening balances, how much history, and the test that decides readiness.
- Go-live night — what actually happens
The order of the night from the inside, the point of no return, and the first week after.
- Why ERP projects fail
Seven reasons, four of them management decisions settled in the first month.
A stalled project, or one not started?
The path serves both: read before signature to build it right, and read mid-stall to work out which station was skipped. Half an hour of diagnostic conversation, at no cost, saves months.
Book a call