Migration Checklist: Moving to GoHighLevel

Phase 0

Decide & Plan

Lock scope, hierarchy, and compliance before any data moves.

• Document the in-scope object list explicitly. Confirm which standard objects (Contacts, Opportunities, Companies) and which Custom Objects are in scope; GoHighLevel caps Custom Objects at 10 per Location, so prune ruthlessly.

  Risk if skipped: Mid-migration discovery that a tenth Custom Object pushes scope past the per-Location cap forces a destructive rethink of the data model.

• Decide the Agency / Sub-account / Location hierarchy up front. Map each operating unit, brand, or tenant in the source to a specific Sub-account/Location in GoHighLevel; this hierarchy controls billing and isolation.

  Risk if skipped: Choosing the wrong granularity forces a rebuild later because Workflows, Custom Fields, and Snapshots are scoped to a single Location.

• Define the historical-data scope in writing. Pick a cut-off date for contacts, opportunities, notes, calls, and emails; decide whether closed-lost opportunities and stale contacts come over at all.

  Risk if skipped: An open-ended historical scope balloons the validation effort by an order of magnitude.

• Decide on the HIPAA add-on before any ePHI moves. GoHighLevel's HIPAA add-on is US$297/month, applies account-wide once enabled, and cannot be disabled; if any Sub-account handles ePHI, enable it before importing.

  Risk if skipped: Loading ePHI into a non-HIPAA account creates a documented compliance breach the moment the import job runs.

• Confirm data-residency constraints with legal. GoHighLevel is hosted on US infrastructure with no EU data residency option; capture sign-off under EU-US Data Privacy Framework or document a blocker.

  Risk if skipped: EU-based contacts loaded without a documented transfer mechanism expose the business to GDPR enforcement after go-live.

• Identify the single accountable migration owner (DRI). Name one person responsible for go/no-go decisions, sign-off on each phase exit, and post-cutover triage.
• Document the success criteria for cutover. Write down what specifically must be true to declare the migration complete: object counts within tolerance, pipeline value reconciled, all Workflows triggering, no failed Bulk Action errors.

  Risk if skipped: Without explicit exit criteria, cutover slips indefinitely while small issues block sign-off.

• Set the cutover window and source-system freeze date. Pick a low-traffic window (often a weekend) and announce the exact UTC start/end times to everyone with write access to the source system.
• List explicitly out-of-scope items. Capture what is not coming over (legacy custom apps, deprecated reports, archived calls before X date) so it cannot reappear as scope creep later.

Phase 1

Pre-Migration Prep

Clean the source, stand up the destination, prep the people.

Phase 2

Source Export

Pull every byte you will need into raw CSVs.

• Export Contacts with every standard and custom field. Include source primary key, email, phone, all custom fields, owner, created date, last modified date, and Tags/labels.

  Risk if skipped: A missed column at export time forces a second freeze window later to extract the missing data.

• Export Opportunities with full field set and pipeline-stage IDs. Include the source pipeline ID, stage ID, monetary value, currency, close date, and the contact/account foreign keys.
• Export Companies/Accounts with primary keys and parent-child relationships. Capture industry, size, address, owner, and contact-to-company association rows.
• Export every Custom Object record with its associated foreign keys. Include the association labels you intend to recreate in GoHighLevel (e.g., Buyer, Account Owner, Assigned Dealer).
• Export Notes as a separate file keyed to source contact ID. Capture note body, author, created date, and the contact/opportunity it attaches to.

  Risk if skipped: Notes that share a file with Contacts lose their author and timestamp once flattened.

• Export Tasks and Appointments as a separate file. Include due date, owner, completion status, and the contact/opportunity association.
• Export call history with recordings and transcriptions if available. Capture call direction, duration, recording URL, transcription text, disposition, and associated contact.
• Export email history (subject, body, direction, timestamps). Email threading does not survive most migrations; export per-message metadata so threads can be reconstructed as Notes if needed.

  Risk if skipped: Lost email history blinds reps to prior conversations and degrades trust in the new system on day one.

• Export the user list with roles, emails, and the source-to-destination owner crosswalk. Owner IDs must map by ID, not by display name; capture both.

  Risk if skipped: Mapping by display name causes silent owner reassignment when two users share a first name.

• Export pipeline definitions and stage order from the source. Capture stage name, position, and probability so Pipelines can be recreated identically before Opportunities load.
• Export attachment file inventory: which files attach to which records. Capture file name, MIME type, byte size, source URL, and the record ID it attaches to; binaries themselves are downloaded separately.
• Download all attachment binaries to a staging location. GoHighLevel CSV uploads cap file size at 100MB on standard plans and 50MB on lower tiers; flag any source attachment that exceeds the cap.

  Risk if skipped: Binaries discovered missing after cutover are usually unrecoverable from a frozen or decommissioned source.

• Screenshot or document every source automation, sequence, and workflow. Workflows do not migrate to GoHighLevel and must be rebuilt by hand; capture triggers, actions, branches, wait steps, and merge fields.

  Risk if skipped: Automations rebuilt from memory weeks later miss edge-case branches that drove revenue or retention.

• Record the UTC timestamp of every export run. This anchors the cutover delta calculation in Phase 5.

Phase 3

Transform & Map

Convert raw source data into GoHighLevel-shaped CSVs.

Phase 4

Sandbox Test Migration

Burn every surprise in the Sandbox, not production.

• Load a representative 1–5% sample of Contacts into the Sandbox via CSV. Use the Bulk Import wizard at Contacts → Import; pick records that exercise every Custom Field, every picklist value, and known edge cases.
• Load a matching sample of Opportunities, Companies, and Custom Objects. Load in dependency order: Companies first, then Contacts, then Opportunities, then Custom Object records, then Associations.

  Risk if skipped: Loading children before parents leaves Associations unresolved and forces a re-run.

• Verify a sample of 10 Sandbox Contacts field-by-field against the source. Open each record and compare every Custom Field, Tag, and Association.

  Risk if skipped: Field-level corruption only surfaces under manual inspection; counts-only validation passes while values are wrong.

• Test edge cases explicitly: long strings, emoji, non-Latin characters, commas in fields. These are the failure modes that pass small-batch testing but fail at production scale.
• Confirm date and timezone arrival is correct on at least 5 Appointment records. Cross-check the displayed time in the destination Sub-account against the source after accounting for the configured timezone.

  Risk if skipped: Off-by-one-day or off-by-hours timestamp errors look fine in spot-checks but corrupt every meeting reminder.

• Verify Associations between Contacts, Companies, Opportunities, and Custom Objects. Open a parent record and confirm related records appear with the correct Association label and cardinality.
• Time each Sandbox load and extrapolate to production volume. If a 5% Contact load takes N minutes, the 100% load will take at least 20N plus throttling pauses; confirm the cutover window is realistic.
• Test attachment upload end-to-end on a representative file. Upload via Media Library, verify the link renders in the contact's File Upload Custom Field, and confirm the file is accessible.
• Rebuild one Workflow in the Sandbox and trigger it with a Sandbox contact. Confirm trigger fires, actions execute in order, and merge fields render correctly.

  Risk if skipped: A Workflow that builds cleanly but fails to trigger is the highest-impact bug to find after cutover.

• Have a domain expert from the using-team spot-check 20 Sandbox records. Sales or marketing leads catch business-meaning errors (wrong stage interpretation, wrong owner) that the migration team cannot.
• Document and resolve every Sandbox issue before declaring Phase 4 complete. Capture each issue in a tracked list with root cause, fix, and re-test outcome.
• Get written sign-off on the Sandbox test results before moving to cutover. Sponsor sign-off is the explicit gate between Phase 4 and Phase 5.

Phase 5

Production Cutover

Execute the sequenced load window with idempotent steps.

• Confirm the source-system freeze is active and no writes occurred in the last hour. Check source audit logs to verify no record was created or updated during the freeze window.

  Risk if skipped: Writes during freeze produce a silent delta the export does not capture, leaving records orphaned in the source.

• Take a final full export of every in-scope object from the source. This is the delta capture between the Phase 2 baseline export and cutover; record the exact UTC timestamp.
• Confirm every destination Workflow, trigger, and integration is paused. Final pre-load check: nothing fires on contact create, tag add, or opportunity stage change.

  Risk if skipped: A single active Workflow can send thousands of unintended SMS or emails during the import.

• Load Users first (or confirm user crosswalk is complete). Owner references in downstream loads resolve to user IDs that must already exist.
• Load Companies before Contacts. Use the dedicated Companies Bulk Import wizard at CRM → Import / Export → Companies; only CSV is accepted.
• Load Contacts via the Contacts Bulk Import wizard. Pick Create / Update / Create + Update mode explicitly; use external ID as the upsert key to make re-runs idempotent.

  Risk if skipped: Using email as the upsert key when duplicates exist silently merges two source records into one destination record.

• Load Opportunities after Contacts. Opportunity rows must reference an existing Contact by primary or secondary key and an existing Pipeline Stage by name.
• Load Custom Object records. Use Bulk Actions → Custom Objects bulk import; pick Create, Update, or Create + Update; configure unique-field dedupe behavior per the mapping spreadsheet.
• Load Associations between objects. Build each Association under Settings → Objects → Associations with the predetermined label and cardinality.

  Risk if skipped: Exceeding the 1,000-record Association cap on a single parent silently truncates the relationship list.

• Load Notes, Tasks, Appointments, and historical activity records. Stamp `source_created_at` on every record; system-managed Created Date will read as cutover day.
• Capture Bulk Action job IDs, row counts, and error logs at every step. Each Bulk Import surfaces a job ID and an error log; export both for the migration audit trail.

  Risk if skipped: Missing job IDs make rollback impossible — Restore Deleted Contacts is the only safety net and it has a retention window.

• Cross-check destination row counts against source row counts after every load. If counts diverge, diagnose before continuing — do not push through with a partial load.
• Re-enable Workflows, integrations, and triggers only after all loads complete. Restart in dependency order: low-traffic Workflows first to confirm clean firing, then high-volume ones.

  Risk if skipped: Re-enabling all Workflows at once can flood A2P 10DLC quota and degrade deliverability for the first day.

Phase 6

Validate

Prove correctness before any end user touches the system.

Phase 7

Post-Migration Cleanup

Close the source, ship the destination, retain the evidence.

• Communicate go-live to all end users with login instructions and a help contact. Include the Sub-account URL, password-reset path, and the first-responder contact for the first 48 hours.
• Run role-segmented training sessions on GoHighLevel basics. Sales reps need Pipelines and Conversations; admins need Workflow builder and Smart Lists; managers need Reporting and Dashboards.
• Set the source system to read-only and announce the read-only date. Revoke write permissions globally; the date triggers the 90-day decommission countdown.
• Rebuild Reports and Dashboards in GoHighLevel from scratch. Source dashboards do not migrate; use the Dashboard builder to recreate the top 5–10 reports the team relies on.

  Risk if skipped: Reports rebuilt only after they're needed delay actionable decisions for weeks.

• Finish rebuilding remaining Workflows beyond the cutover-critical set. Triage by usage volume; the top 20% by trigger frequency typically cover 80% of automation value.
• Re-establish integrations: email/calendar sync, payment gateways, outbound notifications, BI tools. Two-way calendar sync requires each user to re-connect their own Google/Outlook account; the "login as" feature cannot do this.

  Risk if skipped: Users assuming calendar sync "just works" miss meetings until they reconnect their accounts.

• Schedule a 30-day post-migration retrospective. Capture what worked, what failed, and what would be done differently next time.
• Set and document the source-system decommission date. Typically 90 days after cutover, after the GoHighLevel Restore Deleted Contacts window and the 60-day Audit Log retention have both elapsed.
• Archive source-system exports to long-term storage. Keep the Phase 2 and Phase 5 exports in cold storage in case of forensic need; encrypt at rest.

  Risk if skipped: Source data discarded too early leaves no recovery path for disputes that surface months later.

• Export the Bulk Action job logs and Audit Log entries before retention windows close. Audit Logs retain 60 days inside GoHighLevel; the migration evidence must live elsewhere for compliance.
• Close the migration project formally. Final status report, lessons learned, project closure.