The ERPNext post-implementation checklist (60-day playbook for UAE go-lives)

TL;DR

The first 60 days after ERPNext go-live decide whether the project succeeds or quietly rots. Treat it as three phases: hypercare (day 1–14), stabilisation (day 15–45), handover (day 46–60). Track adoption, force the first month-end close on schedule, run a real backup-and-restore test in week 2, sign off role permissions, and refuse to accept legacy-system reports past day 30. The single biggest adoption lever is leadership refusing the old reports. The single biggest data risk is assuming bad pre-migration data fixes itself.

Why this checklist exists

Most ERPNext implementations are scoped, sold, and tracked through to go-live. After cutover, the project plan ends, the partner consultant rotates off, and the client team is suddenly running a system they configured two months ago and a process they have not stress-tested under real load. This is the period where adoption either takes hold or quietly fails — and it is the period that is least often documented.

We have done 400+ ERPNext rollouts since 2017. The pattern repeats: implementations that get the post-go-live phase right keep their users; the ones that drift back to spreadsheets are almost always the ones where nobody was watching the first 60 days. This checklist is what we actually use with clients during hypercare. It is opinionated, sequenced, and written for finance and operations leaders who do not want to be surprised at month-end.

Phase 1 — Hypercare (day 1 to day 14)

The first two weeks are about catching issues fast and protecting morale. Adoption is fragile here; one user who cannot post an invoice on day 3 will tell ten others, and that story lasts a year.

• Daily standups, 15 minutes, fixed time. Partner lead, client product owner, two power users. Topic: what did not work yesterday. Output: a logged ticket per issue with named owner and target date. Cancel the standup the day nobody has anything to raise — not before.
• A single issue tracker, not email. Every defect, configuration gap, training gap, or "it used to do X" goes into one shared list. Email kills traceability and creates parallel realities between partner and client.
• Critical-path watch. Sales invoice posting, payment receipt, stock movement, payroll run. These are the four flows that, if broken, force a fallback to legacy. Confirm each ran cleanly every day for the first two weeks.
• Backup verified by restore test. Day 7 at the latest, restore last night's backup to a staging instance and confirm it boots, opens, and a sample document is intact. A backup that has not been restored is a wish, not a backup.
• Print format sign-off per document type. Tax invoice, credit note, delivery note, quotation, payment receipt, payroll slip. Each one verified bilingual where required and signed off by the document owner. Print formats are the most underestimated source of post-go-live drama.
• Floor-walk in the warehouse and the finance team. Twice in the first two weeks. Watching real users do real work surfaces issues that ticket logs miss — the user who memorised the wrong shortcut, the screen that scrolls badly on the warehouse tablet, the manager who quietly opened the legacy system again.
• Hold the cutover line. No new module additions, no big customisation requests, no "while we are at it" enhancements. Park them on a backlog and revisit after stabilisation. Scope creep in week 2 is fatal.

Phase 2 — Stabilisation (day 15 to day 45)

Hypercare ends when issues stop being daily and start being weekly. The next four weeks are about hardening, closing the first month-end, and starting to migrate trust from the old system to the new one.

• The first month-end close on schedule. Whatever calendar month-end falls inside this window is the close that matters. Do not skip, defer, or run "informationally" alongside legacy. Reconcile bank, AR ageing, AP ageing, inventory valuation, intercompany balances, tax payable. A messy first close is fine. A skipped first close compounds for quarters.
• VAT-201 generated from ERPNext. If a UAE VAT return falls in this window, generate it directly from ERPNext's tax reports. Compare to legacy if you ran parallel — but file from the new system. Filing from the legacy system after go-live signals the team that the old system is still authoritative.
• Data quality scrub on master records. Customer and supplier ledgers reconciled to opening balances. Items with zero stock investigated. Duplicate records merged. The migration loaded what you gave it; the cleanup is your job and ours, jointly.
• Role permissions reviewed and signed off by data owners. Finance head signs off on Accounts module roles. Operations head signs off on Stock and Sales. Document the access matrix in a shared spreadsheet. Default ERPNext role permissions are a starting point, not a finished design.
• Two-factor authentication enabled for all administrators. Non-negotiable. Anyone with System Manager role gets 2FA. Anyone with role permissions to post journal entries should consider it.
• Audit-trail spot check. Pick five Sales Invoices, five Payment Entries, five Journal Entries, five Stock Entries from the past week. Confirm each one shows the correct user, timestamp, and version history in the activity log. If audit trail is broken, find out now, not during a tax inspection.
• Integrations watched closely. Bank statement imports, e-commerce order syncs, courier integrations, payment gateway feeds. First failure is when alerting and retry logic gets stress-tested. Confirm each integration's failure mode produces a notification you can act on.
• Reports rebuilt for the new chart of accounts. Management reports from the legacy system rarely transfer one-to-one. Rebuild the top five used by leadership using ERPNext's Report Builder or Frappe Insights. Get them looking and feeling familiar enough that nobody asks for the old PDF.
• The leadership directive. On day 30, the CEO or CFO sends a written note: "From today, all reports will come from ERPNext. The legacy system is read-only." This single act moves adoption further than any training session. If leadership will not send this note, the implementation has a sponsorship problem that no checklist will fix.

Phase 3 — Handover (day 46 to day 60)

The final two weeks transition from project mode to operating mode. The partner consultant steps back, the AMC contract activates, and the client team carries day-to-day operations.

• AMC (annual maintenance contract) activated. Defined SLA tiers, named contacts on both sides, ticketing channel agreed, escalation path documented. The transition from project comms to AMC comms is a real boundary; mark it explicitly.
• Open-issues triage. Walk the full ticket list with the client product owner. Close everything that is fixed. For everything still open, agree on a category — bug (covered by AMC), enhancement (priced separately), training need (handled by client), or waiting-on-third-party (with named owner).
• Knowledge transfer to the internal product owner. Two structured sessions: one on "how to triage and log a ticket properly", one on "how to make small configuration changes safely" (custom fields, print formats, simple report tweaks). The goal is to make the client autonomous on the small stuff and selective on what reaches the partner.
• Documentation pack delivered. System diagram, integration list with credentials owners, role matrix, backup/restore runbook, escalation contact list. PDF or wiki — somewhere durable, not in someone's email.
• Lessons-learned review. 90 minutes, joint, candid. What went well, what we would do differently, what is still on the watchlist. Output goes into the partner's pattern library and the client's project archive. Skipping this is the project-management equivalent of not eating the leftovers — they go to waste.
• Adoption review. Pull a usage report: documents created per user per week, modules in active use, login frequency. If a module is dark or a user has not logged in for two weeks, find out why now, not at the end of the quarter.

The weekly cadence after day 60

Once the 60-day checklist is closed, the system should be in a steady operating rhythm. We recommend a light recurring cadence:

• Weekly: 30-minute internal review — open tickets, integration health, any user pain points. Add to AMC ticket queue if escalation needed.
• Monthly: Month-end close on schedule. VAT return generated and filed. Backup restore test (yes, every month). One small enhancement shipped if the AMC budget supports it.
• Quarterly: 90-minute partner review. Roadmap, recently-released ERPNext features worth adopting, security review, performance review, AMC consumption review. This is where you decide on the next-tier work — a new module, a new integration, a refresh of a process that has aged.
• Annually: Major version upgrade plan. Tested in a staging environment with regression of critical flows before promotion. AMC contract review and renewal.

The five things that go wrong most often

Pattern recognition from a lot of go-lives. If you only watch for five things in the first 60 days, watch these.

1. Print formats that nobody signed off. Sales invoice missing the supplier TRN footer. Credit note showing the wrong company logo. Salary slip in a format the bank does not accept. Catch in week 2 or watch the entire finance team pretend ERPNext does not work.
2. Opening balances that do not reconcile. Customer ledger sums to a different number than the legacy system. Inventory valuation is off by AED 47,000. Tax payable does not match the last filed VAT return. Reconcile in week 1 or do not, and chase it for six months.
3. Permissions that let everyone do everything. System Manager role distributed too freely during cutover, never trimmed. Anyone can delete a Sales Invoice. Tighten on day 30 or get a tax-inspection question you cannot answer.
4. Integrations that fail silently. Bank feed stops importing on day 12 because the bank rotated an API token. Nobody notices for two weeks because there is no alerting. We see this on roughly one in three go-lives where alerting was scoped out to save a small amount of money.
5. Leadership still accepting legacy-system reports. The biggest adoption killer. If the CEO is fine receiving the old monthly pack from the legacy system on day 35, the message to the team is the new system is optional. Refuse the old reports or accept the slow rot.

How we run the post-go-live phase

Our default contract structure: fixed-fee implementation up to and including 14 days of hypercare, then transition into AMC for the next 11 months. The 60-day checklist above is built into the AMC kickoff — it is not optional, and we do not consider an implementation complete until day 60 is signed off.

We are the only ERPNext Gold Partner in the UAE, which means we have a direct escalation line to Frappe's core team for the rare issue that needs platform help. For UAE-specific items — VAT-201, WPS payroll, designated-zone treatment, multi-emirate consolidation — we have done the work often enough that the post-go-live curve is shorter than for partners who are still learning the local context.

If you have just gone live with another partner and are 30 days in with a checklist that is not working, we also do ERPNext rescue and AMC takeover engagements. Bring the system, the credentials, and the open ticket list — we will run the remainder of the checklist with you.

FAQ