ERP Implementation Timeline: A Month by Month Calendar and Critical Path

10 min read ยท 2026-10-30

A typical erp implementation timeline runs 3 to 4 months for a small business, 6 to 9 months for a midsize company, 9 to 18 months for a large one and 12 to 36 months for a multinational, according to the ranges published by Ultra Consultants. For a midsize company, that usually means planning and design in months 1 to 3, build and data migration in months 3 to 7, testing and training in months 6 to 8, go live at the end of month 9, then hypercare and stabilization through month 12.

This guide is about the calendar: when each phase starts, which phases overlap, which tasks move your go live date, and how the schedule changes with a big bang or a phased rollout. It also lists the timeline questions to ask your vendor before you sign, so the date in a proposal means what you think it means.

The roadmap at a glance

Goal: Go live on a new ERP at the end of month 9 with clean migrated data, tested interfaces and trained users, then close the first month end on time and hand over to internal support by month 12. Duration: 12 months

  1. Planning and design (Months 1 to 3)

    Agree on scope and freeze the design before configuration takes over.

    • Name the sponsor, the project manager and the key users, and agree on scope.
    • Build a baseline schedule and get it signed by the sponsor at the end of month 1.
    • Run process design workshops for every core process in month 2.
    • List every interface and name an owner for each data domain.
    • Close the customization list and freeze the design in month 3.

    Milestone: Design frozen and customization list closed, signed off by the sponsor at the end of month 3.

  2. Build and configuration (Months 3 to 6)

    Configure the system and build customizations and interfaces in short sprints.

    • Start configuration as soon as the design is frozen.
    • Build customizations in sprints, with a key user demo every two weeks.
    • Build and unit test every interface by the end of month 5.
    • Finish configuration in month 6 and run a full process demo with key users.

    Milestone: A full process demo accepted by key users at the end of month 6.

  3. Data migration (Months 3 to 9)

    Run migration alongside the build so clean data is ready for testing and cutover.

    • Take the first data extract in month 3 and give each data owner a cleanup list.
    • Run trial load 1 in month 4 and reconcile it against the old system.
    • Run trial load 2 in month 6 so integration testing uses real data.
    • Run trial load 3 in month 7 and fix every blocking error.
    • Plan the final load and its reconciliation for cutover weekend.

    Milestone: A final trial load with zero blocking errors at the end of month 7.

  4. Testing and training (Months 6 to 8)

    Prove the end to end processes work and get every user ready for go live.

    • Start integration testing in month 6, once interfaces and trial load 2 are in place.
    • Run user acceptance testing in months 7 and 8 with written test scripts.
    • Involve key users during testing, then train end users by role in month 8.
    • Rehearse the cutover with timings, owners and a rollback plan.

    Milestone: A go or no go decision recorded at the end of month 8 against written criteria.

  5. Cutover and go live (Month 9)

    Move to the new system without losing data or a day of operations.

    • Run the final data load and reconcile balances and open orders.
    • Follow the cutover runbook task by task, with an owner and a time for each.
    • Freeze the old system so nobody keeps working in it.
    • Go live at the end of month 9.

    Milestone: The new system live and the old system frozen at the end of month 9.

  6. Hypercare and stabilization (Months 10 to 12)

    Fix issues fast, prove the first close and hand the system to internal support.

    • Review support tickets daily in month 10 with key users and the implementer.
    • Close the first month end in the new system and check it against the old process.
    • Hand over to internal support with documented settings and known issues.
    • Hold a closure review and record the lessons for the next phase.

    Milestone: First month end closed on time in month 10 and the closure report signed in month 12.

How long an ERP implementation takes, by company size and approach

Two published sources give useful planning bands. Ultra Consultants sorts projects by company size. Rand Group, in a guide updated in September 2026, sorts them by approach: a rapid activation close to standard setup, or a full transformation.

Treat these as bands, not promises. Panorama Consulting's 2023 ERP Report found a median project duration of 15.5 months, with only slightly more than half of organizations finishing within their expected schedule, as summarized by VJM Global in June 2026.

Plan at the low end of a band only if you keep standard processes, connect few systems and start with clean data. Otherwise, plan near the top.

  • Small business (Ultra Consultants): 3 to 4 months.
  • Midsize company (Ultra Consultants): 6 to 9 months.
  • Large company (Ultra Consultants): 9 to 18 months, and 12 to 36 months for multinationals.
  • Rapid activation close to standard setup (Rand Group): 6 to 12 weeks for small and midsize businesses, 6 to 18 months for enterprises.
  • Full transformation (Rand Group): 3 to 12 months for small and midsize businesses, 12 to 24 months for enterprises.

ERP implementation timeline month by month: a midsize example

This calendar puts a midsize company at the top of its 6 to 9 month band: go live at the end of month 9, then three months of hypercare and stabilization before the project closes at month 12. Each month ends with one checkpoint you can verify.

Two things this calendar makes visible. Data migration runs from month 3 to month 9, in parallel with the build, not after it. And the project does not end at go live: the first month end close in the new system is the real proof.

  • Month 1: scope, sponsor, project manager and key users named, baseline schedule. Checkpoint: scope and schedule signed by the sponsor.
  • Month 2: process design workshops, interface list, data owners named. Checkpoint: all core processes workshopped.
  • Month 3: design finished, configuration starts, first data extract. Checkpoint: design frozen, customization list closed.
  • Month 4: configuration, customizations built in sprints, trial load 1. Checkpoint: trial load 1 reconciled.
  • Month 5: interfaces built and unit tested, key user demos every two weeks. Checkpoint: all interfaces built.
  • Month 6: configuration finished, trial load 2, integration testing starts. Checkpoint: full process demo accepted by key users.
  • Month 7: integration testing, user acceptance testing starts, trial load 3. Checkpoint: final trial load with zero blocking errors.
  • Month 8: user acceptance testing, end user training by role, cutover rehearsal. Checkpoint: go or no go decision recorded.
  • Month 9: final data load, cutover, go live at month end. Checkpoint: system live, old system frozen.
  • Month 10: hypercare with daily ticket review, first month end close. Checkpoint: first month end closed on time.
  • Months 11 and 12: stabilization, handover to internal support, closure review. Checkpoint: closure report signed.

A 16 week schedule for a small business

VJM Global describes a 16 week plan for a cloud ERP in a smaller company. Its phases overlap by two weeks each, so design starts before planning ends and testing starts before configuration ends.

This fits the 3 to 4 month band from Ultra Consultants. It assumes few interfaces and little customization. Each connected system, such as an online store or a warehouse tool, adds build and test work, so add time to weeks 5 to 14 before you announce a date.

  • Weeks 1 to 4: discovery and planning.
  • Weeks 3 to 6: design.
  • Weeks 5 to 10: configuration.
  • Weeks 9 to 14: testing.
  • Weeks 13 to 16: go live.
  • Week 17 onward: post go live support.

ERP implementation plan timeline: the critical path

The overlaps in the calendar save months, but only if you know which tasks truly block others. These four usually sit on the critical path.

Put these four on one line of your roadmap and review them every week. If one moves, the go live date moves, whatever the other lines say.

  • Design freeze: configuration and interface work depend on it. A late freeze pushes everything after it.
  • Interfaces: integration testing cannot start until they exist.
  • Data cleanup: user acceptance testing on fake data proves little.
  • Key user availability: the same people design, test and train others.

Early signs that the calendar is slipping

Delays rarely show up first in the go live date. They show up in small signals weeks earlier, so check for these at every weekly review and act on the first one you see.

  • Change requests still arrive after the design freeze.
  • A trial data load is skipped or postponed.
  • Key users miss workshops or test sessions.
  • Test scripts are written but not run.
  • Training still has no dates one month before go live.

Big bang or phased rollout: what changes on the calendar

A big bang puts everything live on one date. A phased rollout goes live by module, entity or site, one wave at a time. The choice changes the shape of the whole calendar.

In a phased rollout, draw each wave as its own row on the roadmap: design checks, data migration, testing and go live repeat for every wave. The length of each wave depends on how different the entity or site is from the pilot, so size it after the pilot, not before.

  • Go live: a big bang has one date for everything, a phased rollout has one go live per wave.
  • Total calendar: shorter with a big bang, longer with a phased rollout.
  • Testing: longer and stricter in a big bang, with no second chance, and smaller per wave in a phased rollout.
  • Extra work: a big bang needs a large cutover rehearsal, a phased rollout needs temporary interfaces between the old and new systems.
  • Learning: a big bang trains everyone at once, while lessons from the first wave of a phased rollout shorten the later waves.

Questions to ask your ERP vendor or implementer about the timeline

A date in a proposal means little until you know what it includes. Ask these questions before you sign.

Write the answers into your schedule. Each one turns an assumption into a line you can track.

  • What does your date include? Go live only, or hypercare and the first month end close too?
  • How many hours per week do you need from our key users, and in which months?
  • How many trial data loads are in the plan, and who cleans the source data?
  • Which interfaces are in scope, and who builds and tests each one?
  • What happens to the date if we add a customization after the design freeze?
  • Where is the buffer, and in which phase?
  • Who decides go or no go, and on which written criteria?
  • Can you show the calendar of a similar project you delivered, with its planned and real dates?

How to build your own ERP schedule in five steps

Start from a published band, then adjust it to your own scope. Five steps are enough for a first schedule you can defend.

To present the result to leadership, a one page visual roadmap works better than a long task list. Show the phases, the critical path line and the go or no go date, and keep the detailed task schedule for the project team.

  • Pick your band: use the ranges above, then place yourself low or high in it.
  • Count interfaces and customizations: each one adds build and test time.
  • Plan backward from a realistic go live window: stay away from year end and your peak season.
  • Add a buffer: put it in testing and data migration, where delays usually appear.
  • Add the end of the project: include the go or no go meeting and the stabilization months, so nobody reads go live as the finish line.

Common mistakes to avoid

  • Fixing the go live date before the scope leads to a date nobody can keep, so size the scope first, then set the date.
  • Placing data migration after the build pushes data errors into testing, so start it in month 3, alongside configuration.
  • Treating training as a final week leaves users unprepared, so start with key users during testing.
  • Counting on key users when they have time stalls design and testing, so book their hours in writing with their managers.
  • Ending the plan at go live hides the hardest weeks, so include hypercare and the first month end close.
  • Skipping trial data loads to save time moves errors to go live weekend, so keep every planned load.

Frequently asked questions

How long does an ERP implementation take for a small business?

Ultra Consultants estimates 3 to 4 months. Rand Group puts a rapid activation close to the standard setup at 6 to 12 weeks for small and midsize businesses. Plan for the longer end if your data is messy or you connect several systems.

What is a realistic ERP implementation timeline for a midsize company?

Ultra Consultants gives 6 to 9 months. The calendar in this guide uses the top of that band: go live at the end of month 9, then hypercare and stabilization until month 12.

Is cloud ERP faster to implement than on premise?

Cloud removes server setup and installation, which shortens the start. Design, data migration, testing and training still take the same effort, and most of the calendar sits in those phases.

Why do ERP implementations take longer than expected?

Late customizations, interfaces found too late, dirty data and key users with no time are the usual causes. Panorama Consulting's 2023 ERP Report found that only slightly more than half of organizations finished within their expected schedule. Reviewing the critical path every week catches delays early.

Generate this roadmap with AI