Buying guide - Timeline
How an ERP programme actually runs, phase by phase.
ERP timelines are governed by two things a Gantt chart cannot fix: the condition of your master data and the availability of a safe cutover window. This guide walks through each phase, explains why migration is rehearsed rather than executed once, and sets out the factors that most often move a go-live date.
At a glance
- Who this is for
- Operations and finance leaders planning or reviewing an ERP schedule.
- Key insight
- The first month-end close, not go-live, is the real acceptance test.
- Related
- Cost drivers and ERP migration planning are covered separately.
Phases
Eight phases in an ERP programme.
Migration rehearsals and integration run alongside configuration, not after it.
- 01
Discovery and data assessment
Process mapping, systems inventory, and a hard look at stock accuracy, product data and bills of materials. In ERP this phase determines whether the rest of the plan is realistic.
- 02
Solution design
Chart of accounts, costing method, warehouse and routing structure, approval flows and reporting. Signed off before build, because reversing these decisions later is disruptive.
- 03
Configuration and build
Modules configured to the design, with integrations and any agreed customisation developed alongside.
- 04
Migration rehearsals
Repeated trial loads into a test environment with reconciliation after each. ERP migration is iterative by design, not a single event.
- 05
UAT and process rehearsal
Real people running order-to-cash, procure-to-pay and, where relevant, plan-to-produce end to end against migrated data.
- 06
Training and readiness
Role-based training with the team's own data, plus documented procedures for the tasks that change most.
- 07
Cutover
A sequenced, rehearsed event: stock count, final balances, transaction freeze, load, verify, go. The plan is written and dry-run before the weekend it happens.
- 08
Stabilisation and first close
An agreed post go-live support window through the first month-end. The first close is the real acceptance test for any ERP.
Variables
What decides elapsed time.
| Factor | Effect on elapsed time |
|---|---|
| Master data readiness | The dominant factor. Inaccurate stock and incomplete BOMs turn migration into a data creation project that must finish before cutover can be scheduled. |
| Module scope | Every module adds design, configuration, test scenarios, training and cutover steps. Scope discipline is the most effective schedule control available. |
| Number of sites and entities | Multi-site stock counts and multi-entity balances extend both rehearsal and cutover, and constrain when cutover can happen. |
| Integration dependencies | EDI partners, freight carriers, banks and ecommerce platforms all move at their own pace, and their timelines are outside your control. |
| Financial calendar | Most businesses cut over at a period or year boundary. Missing the window typically defers go-live by a full period. |
| Operational seasonality | Nobody should cut over in peak season. Seasonal businesses effectively have two or three viable windows a year. |
| Customisation volume | Custom development adds build and test cycles, and its defects tend to surface late in UAT when there is least slack. |
| Internal capacity | The people needed for design, testing and cutover are the people running operations. Without backfill, the plan compresses into their spare time. |
Cutover strategy
Phased or big-bang.
Lower risk, longer elapsed
Phased cutover
- Core finance, inventory and purchasing first
- Manufacturing or projects in a later phase
- Smaller change for the team to absorb
- Temporary bridging between phases required
- Value realised earlier on the core
Shorter elapsed, higher risk
Big-bang cutover
- All modules live on one date
- No interim bridging or dual entry
- Very high dependence on rehearsal quality
- Problems are harder to isolate after go-live
- Justified when processes are inseparable
Rather work through your own numbers?
A consultation covers the same ground against your systems, volumes and timeline instead of a general range.
Risk
Five reasons ERP go-live dates move.
- 01
Stock count kept failing reconciliation
Cutover cannot proceed on an unreconciled count. This is the single most common cause of a deferred ERP go-live.
- 02
BOMs were still being written during UAT
Manufacturing testing cannot conclude while the underlying data is changing. It belongs before the test phase, in the plan.
- 03
Costing method changed after design
A late accounting decision ripples through configuration, testing and reporting, and often forces a fresh reconciliation cycle.
- 04
An integration partner was slow
Waiting on an external provider's development or change window is idle time in your plan. Start those conversations in discovery.
- 05
Insufficient rehearsal
A cutover that has not been dry-run will find its problems on the live weekend, when the only options are expensive ones.
ERP timeline questions buyers ask
- How long does an ERP implementation take?
- Longer than a CRM rollout of similar user count, because of migration rehearsals, reconciliation and a controlled cutover. The realistic driver is data readiness and module scope - a phased core rollout with clean data behaves very differently from a full-suite cutover with unreliable stock.
- When is the best time to go live?
- At a period or year boundary, outside peak trading, and when your key operational people are available. In practice this means most businesses have a small number of viable windows each year, which is why cutover timing is planned early rather than negotiated late.
- Can we run the old and new systems in parallel?
- You can, and for high-risk cutovers it is a reasonable control. It is also expensive and tiring, because it means dual entry. If you do it, fix its duration and its exit criteria before you start.
- How many migration rehearsals are needed?
- Enough that a load runs cleanly and reconciles without manual intervention. That is usually several, and the number is an output of data quality rather than a planning input.
- When is an ERP project actually finished?
- After the first clean month-end close in the new system, with balances reconciled and no parallel spreadsheets in use. Go-live is a milestone; the first close is the acceptance test.
Where to go next
Related buying guides
Related architecture and services
Recommended for your industry
Work backwards from a safe cutover window.
Give us your financial calendar and your seasonal peaks, and we will tell you which windows are realistic and what has to be true before each one.