Anyone can draw a methodology slide. What decides an outcome is whether the gates are real, whether the evidence exists, and whether someone is named against every deliverable. This page is our process in the detail we actually run it — for enterprise programmes and for the buildings they run in.
Seven steps. Each one has an input, an output and a decision — and we do not walk through a gate because the calendar says so.
| Step | What happens | What you get | Gate to pass |
|---|---|---|---|
| 1. Align | Working sessions with the executive sponsor and process owners to establish the mission, the measurable objectives and the constraints that are genuinely fixed. | Objectives and success criteria, stakeholder map, constraint register. | Sponsor agrees the definition of success in writing. |
| 2. Assess | Technical, functional and commercial assessment of the current estate: architecture, integrations, data quality, licensing exposure, operational load. | Current‑state assessment, risk register with owners, effort and duration ranges. | No material unknown remains unnamed. |
| 3. Consult | Options presented with trade‑offs, cost shape and risk — including the option to do less, and the option to defer. | Options paper with a recommendation and its reasoning. | Decision made by the client, not assumed by us. |
| 4. Affirm | Scope, sequence, ownership, assumptions and the definition of done are written down and signed before mobilisation. | Charter, RACI, schedule baseline, change‑control procedure. | Every deliverable has a named owner on both sides. |
| 5. Execute | Delivery against the agreed method, with AIDA ECOS™ performing the repeatable build, migration and environment work and recording evidence as it goes. | Working configuration, migrated data, environment parity, weekly evidence pack. | Evidence — not status colour — shows the build meets design intent. |
| 6. Test | Validation against design intent across unit, string, integration, performance and user acceptance, with defects triaged daily rather than banked. | Test evidence, defect burn‑down, performance results, cutover rehearsal outcomes. | Rollback has been rehearsed, not just documented. |
| 7. Deliver | Cutover to a distinct committed date, hypercare with defined response commitments, then transition to run. | Go‑live, hypercare log, run‑book, knowledge transfer, closure report. | The operating team can run it without us in the room. |
This is the loop that runs continuously underneath the phases above. It is the reason our remediation tail is short: nothing waits for a test window to be discovered.
We run the platform vendors' current frameworks rather than a house method. Here is what each phase means in practice, and what the platform contributes to it.
| Phase | Our work | What AIDA ECOS™ contributes |
|---|---|---|
| Discover | Business case, scope shape, readiness assessment and the honest schedule conversation. | Estate inventory and interface discovery from the live systems, so scope is based on what exists rather than what is remembered. |
| Prepare | Charter, governance, team onboarding, environment strategy and the project standards everyone will be held to. | Environment build and refresh automated and repeatable from day one. |
| Explore | Fit‑to‑standard workshops, gap identification, backlog creation and the decisions about where standard process wins. | Configuration baselines captured and compared automatically as decisions land. |
| Realise | Iterative configuration and development, data migration cycles, integration build and progressive testing. | Repeat migration passes on demand; continuous reconciliation and defect surfacing between cycles. |
| Deploy | Cutover planning and rehearsal, dress rehearsal, go/no‑go, production cutover and hypercare. | Machine‑generated cutover evidence and environment verification at each rehearsal. |
| Run | Transition to operations, continuous improvement and the administration of the estate. | Ongoing administration, predictive alerting and the reporting that keeps the estate honest. |
The same discipline maps onto Oracle's unified method and the Oracle Cloud ERP rollout model, and onto the Workday deployment lifecycle — the phase names change, the gates do not.
A facility programme fails quietly in the first six weeks and loudly in the last two. These are the gates where we stop and prove something before money is committed.
Technical and programme authority sits with us; financial authority stays with the owner and the contractor of record. Change orders: we recommend on technical merit, the owner approves cost. The boundary is written into the agreement, not assumed.
Weekly steering with a risk register that names owners and dates, a change log that is actually current, and a decision log so nobody re‑litigates a settled question three months later.
Every critical role has a named alternate agreed before signature. Backups execute the plan — they do not redefine design basis, acceptance criteria or security posture. Absences beyond agreed limits escalate automatically.
Quality acceptance and commissioning never fall back to the party whose work is being inspected. Where certification or tenant audit demands it, the commissioning agent is independent of everyone.
Liquid‑cooling piping crews, water chemistry specialists and rack bring‑up engineers are contracted as a second source at the start, because they cannot be found inside a compressed schedule.
Test results, validation runs, inspection records and commissioning scripts are captured continuously, so the closure pack is assembled rather than reconstructed.
Bad news does not improve with age. A slipping date is raised at the weekly, in writing, with options attached — not discovered at the gate.
Stabilise the environment and protect production first. Root cause analysis follows containment; it does not delay it.
A defect that could recur gets a control — a validation check, a procedure change, a design correction — so the same failure cannot arrive twice.
Every significant incident gets a written debrief that names what happened, including the parts that reflect badly on us. That record is how the next programme gets better.