HubSpot configuration restores require a precise order of operations to preserve data integrity. To restore safely, pause all automations first, then restore items in this sequence: properties, pipelines, permissions and finally workflows. This approach preserves internal IDs and relationships, ensuring that reports and integrations continue to function correctly after recovery.
Key Takeaways
- Order matters: Restore properties first, then pipelines, then permissions, and workflows last.
- Pause first: Always pause active workflows and syncs before beginning a restore to prevent data corruption.
- Preserve IDs: Maintaining original internal IDs is critical for historical reporting and API integrations.
- Validate safely: Test restores in a sandbox environment or on a small data sample before full re-enablement.
Why precise configuration recovery matters
Configuration drift, where actual settings diverge from expected settings, can disrupt revenue operations silently. A deleted pipeline stage can break forecast reports, while a changed property type can invalidate thousands of records.
Recovering these assets is not just about bringing back the data. It is about identity preservation. You must restore the item with its original internal ID and associations intact.
This guide provides a structured runbook that covers how to plan scope, pause automations, run targeted restores and validate results without disrupting production.
Definitions and core concepts
- Configuration drift: Unintended differences between expected and actual settings after changes.
- Identity preservation: Keeping internal IDs, keys and association labels stable during a restore.
- Point in time restore: Recovery to a specific timestamp for the chosen scope.
- Recovery Point Objective (RPO): The maximum acceptable data loss between backups.
- Recovery Time Objective (RTO): The maximum acceptable time to restore operations.
Pre-requisites and guardrails
Before initiating a restore, establish these controls to ensure safety and governance.
- Roles and approvals: Define who requests, approves, implements and validates the change. For production restores, enforce two-person approval.
- Runbook and rollback plan: Keep your documentation current. If the restore fails, you need a defined path to revert changes.
- Pause plan: List every workflow, sales sequence, sync app, import and webhook that must be paused before you begin.
- Evidence capture: Decide where logs, screenshots and sign-offs will be stored for audit purposes.
- Safety environment: Always rehearse complex restores in a HubSpot sandbox first. In production, start with a limited scope to validate behaviour.
Planning the restore: Order of operations
The sequence of your restore dictates its success. Restoring items out of order will break dependencies.
- Properties first: Properties are the containers for your data. You must create or correct definitions and internal names before restoring any item that relies on them.
- Pipelines and stages next: Pipelines depend on properties. Align your stage IDs, required fields and win probabilities. If stage IDs change, historical reporting will fail.
- Permissions and teams: Users and automations need access rights to function. Set owners, teams and permissions before re-enabling any automation that assigns records.
- Workflows last: Workflows act on all the elements above. Restore workflow definitions, enrolment triggers and actions only after the environment is stable.
Mapping dependencies
Before executing the restore, map your dependencies to avoid orphaned references.
- Catalogue workflow references: Check which lists, properties, teams and assets are referenced inside your workflows.
- Record internal IDs: Note the internal IDs for pipelines, stages and properties. These must match the backup for integrations to work.
- Check reporting: Identify any forms, reports or dashboards that rely on the properties or stages in scope.
Scope and method per configuration area
Properties
- What to include: Property definitions, field types, option sets, calculation rules and internal names.
- Failure modes: Renamed internal names prevent data mapping. Type mismatches, such as date vs string, break filters.
- Restore method: Use a point in time property restore. If re-creating manually, use the exact internal name and replicate option keys precisely.
- Validation: Verify that forms load and submit correctly and that workflows resolve property references without warnings.
Pipelines and stages
- What to include: Pipeline definitions for Deals and Tickets, stage IDs, probabilities and required fields.
- Failure modes: Missing stages or changed IDs will break historical reports and forecasting.
- Restore method: Align pipeline and stage IDs to your snapshot. Restore stage order and reapply required fields.
- Validation: Move a test record through each stage. Confirm that required fields enforce data entry and that reports recalculate correctly.
Workflows
- What to include: Enrolment triggers, re-enrolment settings, goals, delays and branching logic.
- Failure modes: Unintended re-triggers can spam contacts. Missing list references can stall branches.
- Restore method: Restore the workflow definition in a paused state. Validate all references.
- Validation: Test enrolment on a small sample list. Confirm each action fires once and ensure no loops occur.
Permissions, users and teams
- What to include: Roles, permission sets, team structure and object access rules.
- Failure modes: Over-permissioning exposes sensitive data. Broken ownership assignments leave records unmanaged.
- Restore method: Restore role definitions and team memberships. Verify access for critical integrations.
- Validation: Perform least-privilege checks. Ensure sample users can complete tasks but cannot access restricted content.
The pause and re-enable sequence
Controlling the flow of automation is critical to preventing a restore storm where automations trigger en masse.
Before restore
- Pause workflows that touch the target objects and properties.
- Pause sales sequences, scheduled imports and active integration syncs.
- Confirm no bulk updates are scheduled during the maintenance window.
After restore
- Re-enable permissions and teams first.
- Re-enable workflows with the lowest blast radius, meaning the lowest impact.
- Monitor for 30 to 60 minutes. Watch enrolment logs, webhook retries and error rates.
- Re-enable high-impact workflows only after validation passes.
Execution checklist
Follow these steps to execute a safe configuration restore.
- Backup health: Confirm backup health for the chosen timestamp. Export a comparison sample.
- Scope definition: Document properties, pipelines and workflows in scope. Note the RPO and RTO.
- Pause: Pause automations and integrations per the pause plan.
- Restore properties: Validate internal names, types and option sets.
- Restore pipelines: Align IDs, stage order and required fields.
- Restore permissions: Verify teams and critical user access.
- Restore workflows: Restore in a paused state. Validate references.
- Logging: Log each action with a timestamp. Capture screenshots of settings.
- Validation: Run acceptance tests. Fix issues before full re-enablement.
- Completion: Announce completion and save the evidence pack.
Validation and evidence
For stakeholders, the evidence of a successful restore is as important as the restore itself.
Acceptance tests:
- Identity: Internal IDs and labels match the snapshot.
- Behaviour: Workflows enrol and act as expected without loops.
- Access: Users have correct access; restricted items remain secure.
Evidence pack:
- Change ticket: Scope, risk assessment and rollback notes.
- Approvals: Two-person confirmation timestamp.
- Restore logs: Actions taken with snapshot IDs.
- Validation results: Checklists and user acceptance notes.
Common scenarios and playbooks
Restoring a Deal pipeline with changed stage IDs
If a pipeline update changes stage IDs, reports will break.
- Identify the correct snapshot and affected reports.
- Pause workflows writing to Deal stages.
- Restore the pipeline configuration to match original IDs.
- Validate by moving a test Deal through stages.
- Recalculate reports to ensure historical continuity.
Recovering from mass property type changes
If a property type change invalidates data, immediate action is required.
- List properties with changed types and dependent assets.
- Pause workflows reading or writing to those properties.
- Restore property definitions to the last good snapshot.
- Validate by sampling records and submitting a test form.
People also ask
How do I preserve internal IDs during a HubSpot restore?
You must use a backup solution that captures configuration metadata. Restore from the exact snapshot containing the original IDs. If you must re-create a property manually, ensure you use the identical internal name and option keys to maintain data mapping.
What should I pause before a configuration restore?
Pause any automation that reads or writes to the affected data. This includes workflows, sales sequences, scheduled imports, integration syncs and webhooks. Resume these only after you have validated the restored configuration.
Can I restore a single pipeline stage without affecting others?
Yes, provided your backup tool supports targeted restores. You must ensure the stage ID and order remain consistent with your reporting history. Always validate workflows that specifically reference the restored stage.
What evidence is required for a configuration audit?
Auditors typically require a change ticket defining the scope, named approvals, restore logs with timestamps and snapshot IDs, and validation results. A summary of lessons learned should be added to the runbook.
