Lesson 19 of 23 · All lessons
Salesforce Platform Administrator II
0 of 23 complete
1Exam Overview
  1. 1.1Platform Administrator II Complete Exam Overview
2Security and Access20%
  1. 2.1Sharing architecture: records, fields, and inherited access
  2. 2.2Profiles, permission sets, delegated administration, and
  3. 2.3Person accounts and complex account-contact models
3Objects and Applications19%
  1. 3.1Object relationships, record types, and schema design
  2. 3.2Applications, Lightning App Builder, Dynamic Forms, and
  3. 3.3Record types, page layouts, and related-list experience
4Auditing and Monitoring10%
  1. 4.1Debug logs and Setup Audit Trail
  2. 4.2Sensitive data, sandbox controls, and security review
  3. 4.3Reviewing security changes and pending access updates
5Cloud Applications11%
  1. 5.1Sales lifecycle: products, price books, schedules, orders, and
  2. 5.2Forecasting and Enterprise Territory Management
  3. 5.3Knowledge, entitlements, and service channels
  4. 5.4Salesforce product suite and platform extension choices
6Data and Analytics Management13%
  1. 6.1Data quality, validation, duplicates, enrichment, and archiving
  2. 6.2Complex reports, dashboards, and analytic design
  3. 6.3External data, backup and restore, and import methods
7Environment Management and Deployment
  1. 7.1Sandboxes, packages, and metadata movement options
  2. 7.2Change set composition, validation, and deployment practice
  3. 7.3Sandbox selection, refresh, and data protection
8Process Automation20%
  1. 8.1Choosing Flow, approvals, formulas, and legacy automation
  2. 8.2Order of execution, recursion, and automation interaction
  3. 8.3Automation troubleshooting, testing, and fault handling

Section 7 · Environment Management and Deployment

7.2 Change set composition, validation, and deployment practice

4 min read Updated Nov 17, 2026

Exam objective

Describe capabilities and best practices for using change sets to move metadata between environments. An outbound change set groups selected metadata components for transfer to a target org with an established deployment connection. The component list should represent a coherent change, not an arbitrary collection. Include dependencies such as custom fields used by flows, record types referenced by layouts, or permission sets needed by a new feature. Salesforce may identify some dependencies, but administrators must still check for missing components and target-specific configuration. A component can exist in both orgs with different values, so understand what the deployment overwrites.

The normal sequence is to develop and test in a sandbox, create the outbound change set, upload it to the connected target, review it there, and validate or deploy. Validation can run tests and check deployment readiness without immediately activating every change, depending on deployment options. Read the result details, warnings, and test outcomes. If deployment fails, identify the missing dependency or incompatible metadata, correct the source or target, and validate again. Avoid repeated production edits that create drift between environments.

Change sets have important limits. They are not a general source repository, do not automatically move records, may not support every metadata type, and do not provide sophisticated branching or full release orchestration. They may not expose the full dependency graph for a complex implementation. Use source control and a release-management tool when the team needs traceability, collaboration, automated comparison, or repeatable multi-org promotion. The exam may ask for an appropriate tool, not a claim that change sets solve every deployment need.

Test planning should reflect the change's risk. A layout change needs a user-interface check; a sharing change needs access tests for owners and nonowners; a flow needs positive, negative, bulk, and recursion-related tests; a permission change needs before-and-after user comparisons. Use a production-like sandbox where appropriate and ensure copied data is protected. Have a named owner verify the deployment in production and observe any asynchronous recalculation or scheduled work.

A rollback plan should be realistic. Some metadata changes can be redeployed in reverse, while data changes or irreversible external side effects may not be easily undone. Preserve the prior metadata version or source commit, document manual steps, and determine how to disable a problematic flow or restore a previous configuration. Communicate a post-deployment validation checklist. A successful deployment status only indicates technical completion, not that users can perform the business process correctly.

Environment-specific configuration deserves special attention. Endpoints, named credentials, email deliverability, connected apps, and integration users may differ between sandbox and production. Do not promote secrets as ordinary metadata or point test automation at real services. Keep a deployment checklist of values that must be set or verified at the target. Coordinate with system owners before changing a connected application or certificate. Exam focus: know the direction of change-set deployment, connected-org requirement, metadata scope, and need for dependency and test review. Distinguish validation from final deployment. Identify when change sets are enough for a simple release and when a source-driven process is better for repeatability or complex dependencies. Keep data migration separate and verify the production user workflow after deployment.

Maintain a short deployment manifest that lists components, dependencies, tests, environment-specific values, approver, and post-release checks.

After deployment, verify more than whether components exist. Confirm the correct flow version is active, page assignments are updated, permission sets are assigned, and automation does not create side effects for unrelated record types. Check one successful case and one rejected or boundary case. Compare a deployment checklist to the actual change set contents and save the validation result for release evidence. If a hotfix is made directly in production, copy that change back to the source environment promptly to prevent later deployments from reversing it.

Practical scenario

A permission-set change enables a new case workflow that depends on a custom field, record type, and active flow. Include and verify all dependencies, validate the change set in the target, and test using a representative support user. After deployment, activate the flow only after checking its version and entry criteria, confirm permission assignments, and verify that external integrations still point to their production endpoints.

Extra concept

Deployment governance is more important as the number of admins grows. A single shared production org can drift when direct changes bypass the release path. Establish who can deploy, how emergency changes are recorded, and how the authoritative source is updated after a hotfix.

Key takeaways

  • Include dependencies and validate in the target before production deployment.
  • Plan for unsupported metadata, environment-specific configuration, and separate data movement.
  • Verify the production workflow after deployment and reconcile emergency changes.
  • Official references: https://help.salesforce.com/s/articleView?id=005298969&language=en_US&type=1 https://help.salesforce.com/s/articleView?id=sf.changesets.htm&type=5 https://help.salesforce.com/s/articleView?id=sf.change_sets_best_practices.htm&type=5 https://help.salesforce.com/s/articleView?id=sf.data_sandbox.htm&type=5
Finished this lesson? Mark it complete to track your progress.

Ready for the real exam?

Take the timed Salesforce Platform Administrator II simulator: same number of questions, same time limit.

Get Complete Access