Written by: Doug Camplejohn, CEO & Co-Founder, Coffee | Last updated: August 8, 2026
Key Takeaways
- Salesforce migrations follow five phases: Discovery, Cleansing, Build, Test, and Cut-over. Each phase expands or contracts based on record volume, integration count, and source data quality.
- Timelines range from 3–6 weeks for under 50K records to 3–9+ months for 1M+ records. Org-to-org and Hyperforce migrations usually add 2–4 weeks.
- Data quality is the biggest risk factor. 91% of CRM data becomes incomplete or duplicated within a year, and fixing issues after go-live costs roughly 100x more than cleaning data before migration.
- 2026 Hyperforce migrations require early Hard Coded URL scans, data residency compliance checks, and production cut-overs scheduled during maintenance windows that can last from a few hours to a full weekend.
- When the core problem is data quality and adoption rather than org architecture, deploy Coffee’s Companion Agent on your existing Salesforce instance to capture and enrich activity data without a migration project.
Five Phases of a Salesforce Migration
- Phase 1 – Discovery & Data Audit (Weeks 1–2): Inventory every object, field, and integration before a single record moves.
- Phase 2 – Data Cleansing & Mapping (Weeks 3–5): Deduplicate, normalize, and document field-level mappings in the source system.
- Phase 3 – Build & Configure (Weeks 6–9): Stand up the target org, disable automation, and execute staged loads in a sandbox.
- Phase 4 – Test & Train (Weeks 10–12): Run UAT, validate reports, and certify users before production.
- Phase 5 – Cut-over & Hyper-care (Weeks 13–16): Execute the production load, monitor data integrity, and maintain the legacy org in read-only mode.
Migration Timelines by Record Volume
| Record Volume | Standard Org Migration (weeks) | Org-to-Org / Hyperforce Migration (weeks) | Recommended Team Size |
|---|---|---|---|
| Under 50K records | 3–6 weeks | 6–8 weeks | 0.5–1 FTE admin |
| 50K–250K records | 6–10 weeks | 8–12 weeks | 1 FTE admin + 1 contractor |
| 250K–1M records | 8–16 weeks | 12–16 weeks | 1–3 FTE |
| 1M+ records | 3–9 months | 4–9+ months | 3–10+ FTE |
Note: activity history volume drives more effort than contact or account counts. A 50K-contact org with 5M activity records requires more work than a 200K-contact org with 500K activities because activity extraction is rate-limited.
Phase 1: Discovery & Data Audit (Weeks 1–2)
Phase 1 creates a clear picture of your current Salesforce landscape before anything moves. You document what exists, how it connects, and who owns each piece.
Deliverables:
- Written dependency map of every object, field, and lookup relationship
- Full integration inventory documenting read/write direction and ownership for each connected system
- Named pilot cohort for sandbox testing
- Rollback baseline: a complete data export via Data Export Service or Data Loader
Resources: 0.5–1 FTE Salesforce admin plus a business stakeholder from sales ops to validate object scope.
Risk call-out: Integration dependencies such as marketing automation syncs, ERP connections, and billing systems are frequently discovered only at go-live. When they are not mapped during week 1, surprises appear late. In one example, a billing system writing renewal dates back to Salesforce caused a customer success team to lose visibility into more than 40 upcoming renewals mid-quarter.
Phase 2: Data Cleansing & Mapping (Weeks 3–5)
Phase 2 focuses on cleaning and standardizing data so you do not move problems into the new org. This work sets the foundation for accurate reporting and adoption.
Deliverables:
The cleansing phase produces four core artifacts that build on each other.
- Field-level mapping document covering every source-to-target pairing
- Deduplication run on email and company domain
- Picklist translation tables resolving value mismatches (for example, “United States” vs. “USA”)
- Removal of tombstone and ghost records such as leads inactive for 3+ years and test records
Resources: 1 FTE admin plus a data steward or contractor for deduplication tooling. Budget 10–20% of total project cost for data cleanup.
Risk call-out: Migrating dirty data instead of cleaning first stretches cleanup that should take two weeks to twelve months, and teams rarely finish that work once new engagement history accumulates in the destination system. Legacy CRM instances often contain test or junk records and duplicate contacts before migration, and 60% of CRM migrations fail due to bad data quality, a problem that follows you to the new org if not addressed here.
Phase 3: Build & Configure (Weeks 6–9)
Phase 3 turns your target org into a safe landing zone for clean data. You configure structure, switch off automation, and run controlled test loads.
Deliverables:
- Target org configured with profiles, permission sets, and custom picklists
- All Record-Triggered Flows, validation rules, and workflow rules disabled on target objects
- Staged sandbox loads in the correct import order: Users/Profiles/Permission Sets, Custom Picklists, Accounts, Contacts, Products/Price Books, Opportunities, OpportunityLineItems, Cases, Activities, Attachments/Files
- Test load of 100 records per object to validate mapping before full execution
Resources: 1–2 FTE admins. Enterprise migrations with multiple source systems benefit from ETL platforms such as Informatica Cloud, Talend, or MuleSoft, while mid-market orgs under 250K records typically use Data Loader with prepared CSV files.
Risk call-out: Automation must be disabled during data loads to prevent firing on every imported record. Leaving it active is one of the most common mistakes that break Salesforce migrations.
2026 Hyperforce Planning Details
Salesforce is retiring the Optimizer app with the Winter ’26 release for all organizations on Hyperforce infrastructure. Non-Hyperforce organizations lose access to all Optimizer features except Hard Coded URL detection after the Spring ’26 release. You can confirm whether you are on Hyperforce by checking your instance name in Setup and comparing it to Salesforce’s instance location documentation.
Teams planning an org-to-org move in 2026 should treat the Hard Coded URL report within Optimizer as a key planning tool. It remains available to support migration planning by identifying fixed references before cut-over. Newer organizations may have limited access to Salesforce Optimizer, so teams should run the Hard Coded URL scan early or use an alternative URL-detection method during Phase 3.
Hyperforce cut-over windows operate as scheduled maintenance events. Production migration cutover takes anywhere from a few hours to an entire weekend depending on data volume. Hyperforce moves also require validation of data residency compliance for any regulated fields before the window opens.
Phase 4: Test & Train (Weeks 10–12)
Phase 4 confirms that data landed correctly and that users can work in the new org with confidence. You validate numbers, fix gaps, and train teams.
Deliverables:
- UAT sign-off from sales, ops, and finance stakeholders
- Post-migration QA checklist completed: record count match per object, 50 child records sampled for relationship integrity, 100 records per object sampled for field population, audit field accuracy reviewed, picklist values checked for case variations
- Pipeline reports reconciled against source system baselines
- Training sessions delivered, and training and change management budgeted at 10–15% of total project cost
Resources: 1 FTE admin plus department leads for UAT. Every dollar spent training an internal Salesforce admin saves roughly three dollars in partner services fees over 24 months.
Risk call-out: Testing of migrated data is frequently cut due to timeline pressure, causing data quality issues to surface only in production. A 2025 migration of 70,000 records completed without UAT produced 12,000 duplicate Accounts and inflated pipeline from $42,000 to $4.2M.
Phase 5: Cut-over & Hyper-care (Weeks 13–16)
Phase 5 moves you into production and stabilizes the new environment. You run the final load, keep systems in sync, and support users closely.
Deliverables:
- Production load executed in batches of 5,000–10,000 records to stay under governor limits
- Legacy org placed in read-only mode, and kept available for 90 days before decommissioning so teams can reference data discrepancies
- Minimum 2-week parallel operation period during which both systems are updated so teams can validate data, integrations, and reports
- Hyper-care support roster with a defined escalation path
Resources: 1 FTE admin on-call plus a partner or senior consultant for the cut-over window. Annual post-go-live operational costs for a 200-user deployment typically range from $90,000–$180,000, including 1 FTE Salesforce admin.
Risk call-out: Ongoing updates in the legacy system during the migration window create sync drift. Delta migrations using timestamp filters and a system freeze during final cut-over are required to reconcile discrepancies.
The five phases above represent the standard path for organizations that need to restructure their Salesforce org. Not every data quality problem requires a migration to solve, so you should confirm the real issue before you commit.
When to Skip Migration and Layer an Agent Instead
A full org migration is the right call when the target architecture genuinely cannot be served by the current Salesforce instance. It is not the right call when the underlying problem is data quality and adoption, and moving to a new org does not fix the process that created bad data in the first place, as noted in Phase 2.
For mid-market RevOps teams whose core problem is low adoption, missing activity data, and inaccurate forecasts, a Companion Agent layered on top of the existing Salesforce instance resolves the data quality problem without the cost, timeline, or change-management burden of a full migration. Coffee’s Companion App deploys an autonomous agent that captures emails, calendar events, and call transcripts, then writes structured data back to Salesforce automatically. The system of record stays clean while reps avoid acting as data entry clerks.
Migrate vs. Companion Agent Comparison
| Dimension | Full Org Migration | Coffee Companion Agent on Existing Salesforce |
|---|---|---|
| Timeline | 6–16 weeks for mid-market | Days to authenticate and begin writing data back to Salesforce |
| Cost (services) | $40,000–$75,000 for mid-market with 3–4 integrations | Seat-based subscription, with no migration services fee |
| Data quality outcome | Dependent on pre-migration cleansing, and only 46% of migrations finish on time and 64% run over budget | Agent continuously captures and enriches records from emails, calendars, and transcripts going forward |
| Change-management effort | 5–10 hours of rep learning time per person plus a parallel licensing period | Reps keep existing Salesforce workflows, and the agent handles data entry in the background |
15-Item Go-Live Checklist
- Full data export completed and stored as rollback baseline
- All automation (Flows, Process Builder, Apex triggers) disabled on target objects
- Validation rules audited and selectively disabled for migration user
- Field mapping document finalized and signed off by data owner
- Duplicate matching rules configured before first import
- External IDs staged on all parent objects
- 100-record sandbox test load completed per object
- Import order confirmed: Accounts, Contacts, Opportunities, Activities, Attachments
- Batch sizes set to 5,000–10,000 records per load
- Integration dependency audit completed with read/write direction documented
- UAT sign-off obtained from sales, ops, and finance
- Pipeline reports reconciled against source system baselines
- Cut-over window scheduled for off-peak hours or weekend
- Legacy org set to read-only mode post cut-over
- Hyper-care support roster active for minimum 2 weeks post go-live
If the checklist reveals that the primary gaps are data quality and rep adoption rather than org architecture, the Companion Agent path eliminates steps 1–13 entirely. Get started with Coffee and deploy the agent on your existing Salesforce instance without a migration project.
Frequently Asked Questions
How long does a Salesforce org-to-org migration actually take for a 100-person sales team?
A 100-person sales team running a mid-complexity Salesforce org, with custom objects, Flow automation, and three to five integrations, should plan for 10 to 14 weeks from kickoff to go-live. The data cleansing phase is the most variable. A moderately messy source org adds one to two weeks before migration can begin, while a heavily customized org with undocumented automation can add four to six weeks. Teams that skip the data audit and attempt a compressed timeline routinely encounter duplicate records, orphaned contacts, and inflated pipeline figures that require months of post-launch remediation.
What is the biggest risk factor in a Salesforce migration timeline?
Source data quality is the single largest variable. 91% of CRM data becomes incomplete, stale, or duplicated within a year, and the relative cost of fixing data quality is approximately 1x before migration, 10x during migration, and 100x after go-live. The second-largest risk is undiscovered integration dependencies. Billing systems, marketing automation syncs, and ERP connections that write data back to Salesforce are frequently identified only at cut-over, not during planning. A full integration dependency audit in week one, documenting every connected system’s read/write direction and ownership, is the most effective single mitigation step.
What are the 2026 Hyperforce migration requirements I need to plan for?
Salesforce is retiring the Optimizer app with the Winter ’26 release for all Hyperforce organizations, and non-Hyperforce organizations lose access to all Optimizer features except Hard Coded URL detection after the Spring ’26 release. Newer organizations may have limited access to Salesforce Optimizer. Teams planning a Hyperforce org move in 2026 must run the Hard Coded URL scan early to identify fixed references, validate data residency compliance for regulated fields before the cut-over window, and schedule the production load during a maintenance window that can span a few hours to a full weekend depending on volume. Confirm your instance’s Hyperforce status in Setup before scoping the project.
Is there an alternative to a full Salesforce migration if the core problem is bad data and low adoption?
A Companion Agent layered on the existing Salesforce instance addresses data quality and adoption without a migration project. The agent captures emails, calendar events, and call transcripts autonomously, then writes structured activity data, enriched contact records, and deal updates back to Salesforce in real time. Reps retain their existing workflows and the system of record stays unchanged.
This approach is appropriate when the org architecture is sound but the data entry process is broken, which describes many mid-market Salesforce deployments where 71% of sales reps report spending too much time on manual data entry. A full migration is appropriate when the org structure itself must change due to a merger, Hyperforce deadline, or multi-cloud consolidation.
For RevOps leaders who need accurate pipeline data and forecasts without a 12-week migration project and a $40,000–$75,000 services engagement, the Companion Agent is the lower-risk path. Get started with Coffee and have the agent writing clean data back to your Salesforce org within days.
