Deploying changes from a Salesforce sandbox to production is a critical milestone in any Salesforce project. A rushed or incomplete go-live can lead to broken automation, security issues, and unhappy users. Whether you’re a Salesforce Admin, Developer, or Technical Lead, having a structured go-live checklist ensures a smooth, predictable release.
This blog walks you through a practical, real-world Salesforce go-live checklist that covers everything from testing and security to deployment and post-production monitoring.
1. Pre-Deployment Readiness
Before anything is deployed, the sandbox must be fully stable and signed off.
Functional Readiness
Ensure all requirements have been implemented and validated by the business.
User interfaces (Lightning pages, LWC components, Flow screens), automation behavior, and any Agentforce or AI-driven actions should be reviewed and approved.
Testing & Code Quality
- Code coverage is 75%
- Remove all deadcode
- Manifest items list - All metadata or manual configuration is listed
- Tested and Business Signed-off functionality.
- Hard-coded URL :- Sometimes we hardcode URL or there are scenarios which are specific to org. Make sure that you document all such scenarios in manual deployment steps or post deployment steps
Salesforce enforces a minimum of 75% Apex code coverage, but passing this threshold alone isn’t enough. All test classes must pass consistently, and the code should be bulk-safe, governor-limit friendly, and free of hard-coded IDs or URLs. Debug statements should be removed, and SOQL queries should be optimized. Always test bulk scenarios and large data volumes.
Configuration Validation
Confirm that custom objects, fields, validation rules, picklists, page layouts, and Lightning record pages are finalized. Any configuration that could block production data—such as strict validation rules—should be carefully reviewed.
2. Security and Access Review
Security issues are one of the most common causes of post-go-live failures.
Profiles and Permission Sets
Verify that all required permission sets exist and grant correct object, field, Apex class, and Flow access. If Agentforce or custom automation is involved, ensure the correct permissions are assigned to the right user groups.
Never rely on System Administrator testing alone. Validate access using real end-user profiles.
Sharing and Visibility
Review org-wide defaults, sharing rules, and role hierarchy behavior. Test access using real user profiles to confirm users can see and edit exactly what they should—no more, no less.
3. Data and Metadata Preparation
A perfect deployment can still fail if the underlying data isn’t ready.
Data Readiness
Make sure essential master data—such as Accounts, Products, Price Books, and Record Types—exists in production. Sandbox-only test data should not impact production processes. Data quality checks, including duplicate management, should be completed in advance. Activate records wherever required.
Metadata Dependencies
Confirm all dependencies are included in the deployment package. This often includes custom labels, custom metadata types, named credentials, remote site settings, and custom settings. Missing dependencies are a common cause of deployment failures.
4. Integrations and External Systems
Integrations require extra attention during go-live.
Production endpoints, credentials, and authentication mechanisms must be configured correctly. Named credentials should be tested, and error-handling logic verified. If your solution sends emails or triggers external systems, validate that these actions behave correctly in production.
5. Deployment Execution
A successful go-live depends on a controlled deployment process.
First, confirm the deployment approach—Change Sets, Salesforce CLI, or a CI/CD pipeline—and get approval for the deployment window. During deployment,
- move all required metadata,
- resolve any errors,
- assign permission sets,
- activate Flows and approval processes,
- schedule batch or scheduled jobs if applicable.
- Assign page layouts and activate lightning pages
6. Post-Deployment Validation
Once the deployment is complete, immediate validation is essential.
Smoke Testing
Log in as an end user and perform basic actions such as creating and editing records, triggering automation, and verifying emails. Reports and dashboards should also be checked to ensure data is displaying correctly.
Automation and Agentforce Validation
If your org uses Agentforce or advanced automation, validate responses, invocable Apex methods, and error handling. Temporary logging can help identify issues early.
7. User Enablement and Communication
Technology alone doesn’t make a go-live successful—users do.
Provide training, share documentation, and clearly communicate what has changed. A go-live announcement should include known limitations, support contacts, and expectations for the first few days after release.
8. Monitoring and Hypercare
The first few days after go-live are critical.
Monitor debug logs, error emails, and integration failures closely. Establish a short hypercare period—typically 24 to 72 hours—during which issues are prioritized and resolved quickly.
9. Rollback and Backup Strategy
Even the best deployments can encounter issues.
Ensure you have a documented rollback plan, recent data backups, and access to previous metadata versions. Emergency access should be available in case immediate fixes are required.
10. Final Sign-Off
The go-live process isn’t complete until formal sign-off is received. Business stakeholders and technical teams should confirm that the release is stable, functional, and ready for full use.
Conclusion
A Salesforce go-live doesn’t have to be stressful. With a structured checklist and disciplined execution, you can move from sandbox to production confidently and minimize risk. Treat go-live as a process—not an event—and you’ll set your Salesforce org up for long-term success.
.png)
