Octagotech · Certified Odoo Ready partner · octagotech.com · sales@octagotech.com · USA +1 917 254 9416 · India +91 8153897751

Guide · Printable checklist

The 12-point Odoo pre-go-live checklist

By Octagotech TeamUpdated

The checklist our project managers sign off before any Odoo deployment goes live. It does not cover every project's specifics, but it covers the checks that are easiest to skip under deadline pressure.

Book a discovery call

Save as PDF opens your browser's print dialog. Your ticks print with the page.

12checks6groupsFreeno email gate

This guide

What is inside

Checks
12, in six groups
Use it
In the final weeks before cutover
Time to read
About 7 minutes
Format
Web page that prints to PDF
Your ticks
Saved in this browser only
Cost
Free, no email gate

0 of 12 checked · ticks stay in this browser

Checks 1–2

Data and migration readiness

0 of 2
  1. Run the full migration at least twice: once early to surface structural issues, and a final run on a near-production dataset no more than a week before go-live. A business stakeholder, not only the IT lead, reviews the migrated records and signs off.

    Skip it and: day-one surprises, like open purchase orders with the wrong supplier or customer balances that do not match your AR aging report.

  2. Test with your real product catalog, customer accounts and edge cases: the three-way match that breaks on a partly received PO line, payroll for employees with mid-month start dates, a manufacturing order with nested sub-assemblies. Every user should be able to repeat their hardest weekly task in staging.

    Skip it and: false confidence. Hearing "this worked in testing" on day one usually means a simplified version of the workflow was tested, not the real one.

Checks 3–4

Integrations and infrastructure

0 of 2
  1. Re-test every integration against live endpoints: payment gateway, shipping carrier API, EDI connection, third-party WMS, accounting sync. Staging almost always runs on sandbox keys, test webhooks and dummy credentials.

    Skip it and: token expiry, IP whitelisting and rate limits behave differently in production. For example, a staging Stripe connection can work perfectly while the live key has a webhook URL mismatch that nobody catches until the first real sale.

  2. A step-by-step document with timestamps, named owners and explicit go/no-go points. It answers three questions: what is the last moment you can abort and revert to the legacy system, who makes that call, and how long a rollback takes, including what data entered in Odoo would be lost.

    Skip it and: a failed cutover has no way back. Teams with a tested rollback path make cleaner go/no-go decisions under pressure.

Checks 5–6

Security and financial configuration

0 of 2
  1. IT sets the roles up; only department heads know whether they match real jobs. Each head logs in with a test account at their permission level and tries three things they should not be able to do. A warehouse picker should not see vendor pricing, and an accounts payable clerk should not post journal entries directly.

    Skip it and: Odoo's permission model is granular, which makes it easy to misconfigure. If any of the three attempts succeeds, the role is wrong.

  2. Every legacy account maps cleanly to the new structure, and opening balances are entered and verified before the first real transaction posts. Run a trial balance in both systems for the same period and reconcile the difference to zero, or document any intentional adjustment.

    Skip it and: finding a discrepancy in retained earnings two weeks after go-live is a bad day for everyone.

Checks 7–8

Operations and automation

0 of 2
  1. Send a test email from every major document type (sales order, invoice, purchase order, delivery) and confirm it arrives correctly formatted in an external inbox. Check that inbound aliases for support tickets or sales inquiries route correctly.

    Skip it and: order confirmations, supplier purchase orders, invoice reminders and picking notifications all depend on a correctly configured outbound mail server.

  2. Open the Scheduled Actions list and audit every enabled action. Confirm you want it running, know its frequency, and know what its first run will do to live data.

    Skip it and: payment follow-ups, replenishment rules, subscription renewals, payroll batches and EDI syncs fire as soon as you are in production. An overdue-payment reminder at 2 AM on day one, before your AR team is trained, is predictable and easy to prevent.

Checks 9–10

Reporting and training

0 of 2
  1. Pick five to ten reports the business relies on weekly, such as inventory valuation, sales by product category, aged receivables, payroll summary and production efficiency. Run them in both systems for the same period. The numbers do not have to match exactly, but every variance must be explainable.

    Skip it and: you cannot trust reports you cannot explain, and misconfigured accounting settings wait until month-end close to show up.

  2. Name one power user per department who can run the end-to-end workflow without help from the implementation team. Warehouse: receive a purchase order, put it away, and pick against a sales order. Accounting: enter a vendor bill, match it to a PO, and run a payment batch.

    Skip it and: if your power users cannot do it alone, the wider team cannot either, and the day-one support load becomes unsustainable.

Checks 11–12

Support and recovery

0 of 2
  1. Every user knows who to call and what "critical" means in response time. Tier 1, how-to questions, goes to your internal power user or help desk. Tier 2, configuration issues, goes to your implementation partner. Tier 3, data corruption or system down, is an emergency escalation with a defined response SLA.

    Skip it and: the escalation path only works if it was communicated and tested before the pressure is on. At Octagotech, critical issues on the Enterprise tier reach an on-call engineer 24/7.

  2. Run a full restore drill: restore the most recent backup to a clean environment, confirm the application comes up with all data intact, and time the process. On Odoo.sh this is straightforward; self-hosted instances need a more deliberate process.

    Skip it and: "we have backups" and "we can restore in under two hours" are different statements. If a restore takes six hours, your contingency plan has to change.

Go / no-go

Sign-off

A signed checklist turns assumptions into documented facts. Fill it in on screen, or print it and sign by hand.

Why it matters

Each check takes a minute to skip and days to recover from

A missed data sign-off leads to a week of manual corrections. An untested rollback leaves a botched cutover with no way back. Undertrained users swamp the internal team and the implementation partner at the same time.

Experienced teams do not forget these steps; project pressure compresses timelines and invites optimistic assumptions. The full article walks through each point: the 12-point Odoo pre-go-live checklist. For how we structure a project so go-live is less stressful, see Odoo implementation services.

Going live soon?

We are happy to do a pre-launch review of your readiness, or talk through where your project stands in a 30-minute call.

Book a discovery call

Get a second set of eyes before go-live

The first call takes 30 minutes.

Or email sales@octagotech.com