A Practical Salesforce Release Testing Checklist

Salesforce upgrades your org three times a year and your team ships changes in between. This checklist keeps both kinds of release from surprising your users.

Type
Article
Topic
CRM & Engineering
Reading time
5 min
Updated
Author
GeekMidst Team
Reviewed by
Site owner
Banner for the GeekMidst article “A Practical Salesforce Release Testing Checklist”, covering Before you start and Seasonal release routine

Two kinds of release change a Salesforce org. Salesforce itself ships three major releases a year (Spring, Summer and Winter), applied to your org whether or not you are ready. And your own team deploys configuration and code on its own schedule. Both can break a flow, a page layout or an integration, and both deserve a repeatable testing routine rather than a scramble the week before.

This checklist is written for platform owners and small delivery teams who do not have a dedicated QA department but need to release with confidence.

Checklist of six steps for testing a Salesforce seasonal release: release notes, preview sandbox, automated tests, critical processes, triage and user preparation
Read the notes early, test in a preview sandbox against a baseline, and have process owners sign off before the production upgrade date. View full size

Before you start

You need three things in place before any checklist helps:

  • An inventory of what matters. A short list of business-critical processes: lead capture, opportunity to order, case handling, marketing sync, key integrations. If it is not on the list, it will not be tested.
  • A sandbox strategy. At least one sandbox that mirrors production configuration closely, refreshed on a known cadence. For seasonal releases, Salesforce upgrades some sandbox instances ahead of production during the preview window; check the release dates and your sandbox instance on the Salesforce Trust site and in the release notes.
  • Named testers. Business users who own each critical process and will sign off on it.

Seasonal release routine

  1. 01

    Read the release notes early

    When release notes are published, filter for features you use and for changes marked as enforced or auto-enabled. Note anything touching security, authentication, Flow, Lightning components or APIs your integrations call.

  2. 02

    Confirm a preview sandbox

    Check whether your chosen sandbox will be on the preview instance. If not, refresh or create one in time to be upgraded with the preview.

  3. 03

    Run the automated tests

    Run all local Apex tests and any UI or integration test suites in the upgraded sandbox. Compare results with a run from before the upgrade.

  4. 04

    Walk the critical processes

    Have process owners run their scripted scenarios end to end, including integrations and email or notification steps.

  5. 05

    Triage and log

    Record every difference, decide whether it is a defect in your org, a Salesforce change you must adapt to, or a known issue, and assign an owner.

  6. 06

    Prepare users

    Tell users about visible changes before the production upgrade date, and plan any configuration changes you need on or after that date.

What to test, by risk

Why it breaksHow to test
Apex and triggersAPI version behaviour, test data assumptionsRun all local tests and compare with a baseline
FlowsChanged element behaviour, new validationScripted scenarios with bulk and single records
Custom LWC and AuraPlatform changes to base components and securityLoad each page that hosts them, check console errors
IntegrationsAuthentication, API changes, IP or TLS changesRun a full sync cycle and check error stores
PermissionsNewly enforced permission changesLog in as each key profile or permission set group
Reports and dashboardsField or object changesOpen the executive and operational dashboards

Checklist for your own releases

Seasonal releases are predictable. Your own releases are where most incidents start, because the change is new and the person who built it is often the one who tested it.

Scope and build

  1. Every change is linked to a user story with acceptance criteria someone other than the builder can check.
  2. The release contents are listed as components (or a source-tracked branch), not as "everything Priya worked on".
  3. Dependencies are identified: fields referenced by flows, permission sets that grant new fields, picklist values used in automation.

Test

  1. Each story is tested in a sandbox by someone who did not build it.
  2. Automation is tested with bulk data, not just one record. A flow that works on one record can fail on two hundred loaded by Data Loader or an integration.
  3. Apex tests pass with meaningful assertions, not just coverage. See writing reliable Apex tests.
  4. A regression pack covering the critical process list is run, even if the change looks unrelated.
  5. Users with different profiles test the screens they actually use.

Release

  1. The deployment is validated against production before the release window, with tests run.
  2. A rollback plan exists for each component: what to redeploy, what to deactivate, and what data might need correcting. Some changes, such as deleted fields and changed data, cannot simply be rolled back.
  3. Data changes (backfills, picklist migrations) have their own script and their own check.
  4. Users are told what changed and where to report problems.

After release

  1. Smoke test the critical processes in production within the release window.
  2. Watch error stores, flow error emails and integration logs for at least a day.
  3. Hold a short review: what slipped through and which check would have caught it.

Common mistakes

  • Testing only in a developer sandbox. Developer sandboxes do not contain production data, so data-dependent logic and sharing behaviour may not show up. Use a partial or full copy sandbox for regression where you can.
  • Testing as a System Administrator. Admins bypass many permission problems that real users will hit.
  • Skipping integration tests because the integration "has not changed". The thing it talks to may have.
  • No baseline. Without a record of test results before the change, you cannot tell whether a failure is new.
  • Release notes read by one person. Share a short summary with the team and process owners.

Keep the checklist alive

A checklist that never changes is a sign nobody uses it. After each release review, add the check that would have caught the issue, and remove checks that never find anything and cost real time. Store the checklist next to the release plan so it is used, not filed.

Want a release process your team can repeat?

We can help you set up regression packs, sandbox routines and sign-off steps that fit the size of your team.

Talk to a Salesforce expert →

Tell us what you are trying to fix.

Share your Salesforce setup, goals and constraints. A practitioner reads every message and replies with questions, first thoughts or an honest view of whether you need outside help at all.

Talk to a Salesforce Expert
Book a Consultation