A Go-Live Checklist for Custom Business Software: Data, Training and Ownership
A system can pass a demonstration and still be unprepared for everyday business use. Real operations bring incomplete records, forgotten passwords, staff absences, unusual transactions and customers who do not follow the expected path.
A go-live checklist forces the team to confirm not just that features exist, but that the organisation can run the process and recover when something goes wrong.
Agree acceptance criteria before launch
List the key workflows and what success looks like for each one. Confirm that the right people have tested them using realistic scenarios. Record defects, assign severity and decide what happens to each.
What must be written down, not assumed
“Looks good” is not a reliable acceptance record.
- Must fixBlocks a priority workflow. Do not launch without it.
- WorkaroundHas an agreed workaround with a named owner.
- Can waitLower impact. Accepted deliberately, not ignored.
Validate data and opening balances
Identify what will be imported or entered at launch, who checked its accuracy and how duplicates or incomplete records will be handled. If the system begins with balances, active bookings, open jobs or current customer records, reconcile them against the agreed source.
Confirm access and daily ownership
Make sure each staff role works with the intended permissions, administrators are named and support contacts are available.
Tasks that need a named person
Avoid relying on an unnamed “IT person” to own every operational decision.
Rehearse the first day and a fallback
Walk through the first real operating day with the people who will use the system. Explain where to report errors, what not to retry and how to handle time critical work if the system is unavailable.
- Who does what, in order
- Where errors are reported
- What must not be retried
- How urgent work is prioritised
- How work continues if the system is down
- Who records the transactions
- How they are entered later
- How duplicates are prevented
Define a rollback or pause threshold
Before launch, decide which failure would justify pausing the rollout. Identify who can make the decision and what the temporary operating procedure will be.
Failures that may justify pausing
- Essential records cannot be accessed
- Billing produces duplicate charges
- A critical workflow cannot be completed
- Data integrity cannot be trusted
A rollback plan is not an expectation of failure. It is a way to respond deliberately if a serious issue occurs.
Secure the starting state
Before importing data or switching systems, confirm the appropriate backup or export exists and that authorised staff know how it will be accessed. Record the time at which it was taken.
- Step 1Confirm the backup exists
- Step 2Record when it was taken
- Step 3Protect the copy
- Step 4Plan for post backup changes
Give staff a short launch reference
A one page guide should tell users how to sign in, complete common tasks, identify a failed action and request support. Keep detailed technical documentation available for administrators.
What staff actually need on day one
Do not expect every employee to read a long manual before starting work. Ensure that the guide reflects the actual version and process being launched.
Schedule post launch reviews
Hold short reviews after launch to gather feedback, check key data and inspect unresolved issues. Confirm who owns fixes and how changes are prioritised.
Review the launch decision with evidence
At the readiness meeting, review the evidence and record accepted risks along with the person who accepted them.
What the evidence should cover
If a launch item is incomplete, decide whether the gap is safe to defer or whether it must be resolved first.
Run the go live review with evidence
List critical workflows and confirm the agreed tests have passed. Record unresolved defects, accepted risks, workarounds and the person who authorised each decision. Critical failures should not be waved through because a launch date is close.
Validate imported records, roles, support contacts, user training and backups. Rehearse how to pause or roll back if a serious problem occurs and explain the fallback process to staff.
Schedule short post launch reviews, including a check on errors, user questions and data reconciliation. Keep a prioritised issue list and assign owners. Close the launch phase only when the organisation can complete daily work, respond to failures and maintain the system without relying on informal knowledge held by one person.
“Ready” means the priority workflows have passed agreed tests, critical defects are resolved, initial data is checked, roles and support contacts are known, and a fallback has been rehearsed. If one of these is missing, document the risk and make a deliberate decision rather than hoping it will not occur.
Frequently asked questions
Who should approve go live?
The owner or authorised project lead should review test evidence, data readiness, support, risks and fallback arrangements.
Should unresolved defects always block launch?
Critical defects should normally block launch; lower risk issues may be accepted deliberately with an owner and workaround.
Action checklist
- Acceptance criteria and test results are documented.
- Launch data is validated and reconciled.
- Roles, owners and support routes are clear.
- Fallback handling has been rehearsed.
- Post launch reviews and change control are scheduled.
Final thought
A disciplined launch protects the investment and gives staff a clear way to work through unexpected issues. Make readiness measurable, assign each task and keep a short review cycle after go live.
Preparing a system for launch?
Stonehenge Digital KE's Business Management System service includes training and handover and can be discussed alongside the project's acceptance needs.
Explore the Business Management System Service →Interested in Business Management System?
Let's build this for your business
Replace spreadsheets, paper records, and WhatsApp chaos with a custom system built for exactly how your business works.