Orbital / CRM data migration checklist for New York service businesses

CRM data migration checklist for New York service businesses.

Plan a safer CRM migration with data inventory, field mapping, cleanup, access controls, test imports, cutover, validation, rollback and vendor handoff.

CRM migration · New York service-business guide

By Orbital Content Team · Published November 4, 2026

AI-assisted research and drafting with editorial review. Current FTC, NIST and New York Attorney General guidance was checked on November 4, 2026. The migration workbook, testing sequence, cutover plan and acceptance framework are Orbital editorial recommendations. Orbital did not access a business's systems, inspect or migrate customer records, test a CRM, determine legal obligations, estimate a project or claim client results.

A CRM migration is not a file-upload task. It is a controlled decision about which records a business will carry forward, how old fields will map to the new operating model, who may see or change the data and what evidence will prove the new system is trustworthy.

For a New York service business, customer data may arrive through forms, calls, scheduling systems, email, spreadsheets, invoicing tools and an existing CRM. Copying every historical record into a new platform can preserve duplicates, stale permissions, unclear consent, broken ownership and reporting definitions that the project was supposed to fix.

This checklist starts before an export is created. It produces a decision brief, migration workbook, test record, cutover plan and evidence pack that can be used with an internal administrator, CRM vendor or implementation consultant.

1. Write the migration decision brief

Begin with the business reason for moving. A migration may support a new customer workflow, replace unsupported software, consolidate spreadsheets, connect marketing with sales, improve service handoffs or create clearer owner reporting. State the decision in operational terms rather than naming features.

Questions to settle before selecting the migration method
DecisionRecord in the briefWarning sign
Business outcomeThe customer or staff action that should become clearer, faster or more reliable.Success is defined only as “all records imported.”
ScopeIncluded systems, record types, date range, attachments, activities and users.The provider assumes every export column and file should move.
OwnersExecutive sponsor, process owner, data owner, technical lead, approver and vendor contact.No one can approve exclusions, duplicate rules or final acceptance.
ConstraintsBlackout dates, contractual requirements, retention rules, integrations and staff availability.Cutover is scheduled before constraints are documented.
AcceptanceCounts, samples, workflows, permissions, reports and reconciliations that must pass.The only test is whether the new CRM opens.
FallbackHow work continues if the import, integration or cutover fails.The legacy system will be disabled before a recoverable checkpoint exists.

Keep the migration separate from unrelated customization. A new pipeline, lead-scoring model, automation program and data migration can influence one another, but bundling every change into one irreversible launch makes defects harder to isolate. Define the minimum operating model required for cutover and move lower-priority enhancements into a controlled backlog.

Use Orbital's CRM implementation checklist when the business still needs to define workflows, requirements, ownership and rollout before it can plan the migration in detail.

2. Inventory systems, data and owners

Create a source-of-truth register before exporting anything. The FTC's current small-business cybersecurity guidance recommends maintaining an inventory of hardware, software, data and services. For a migration, that means tracing where customer information is collected, changed, copied, reported and retained.

Minimum source-of-truth register
System or fileDocumentDecision
Current CRMObjects, fields, record counts, owners, activities, files, automations and exports available.Authoritative source for each record type and the point at which changes freeze.
Forms and call systemsCaptured fields, timestamps, attribution, consent language, routing and failed deliveries.What history moves and how new inquiries reach the target CRM during cutover.
Email and schedulingContacts, events, messages, permissions and synchronization rules.Which history is necessary and whether links or summaries are enough.
SpreadsheetsOwner, purpose, last updated date, formulas, duplicates and hidden columns.Merge, archive, correct or exclude.
Billing and delivery toolsCustomer identifiers, matter or project references, status, financial fields and integration direction.Which system remains authoritative and which values are display-only.
Analytics and advertisingLead identifiers, source labels, campaign fields and retention limits.Minimum attribution history needed for useful reporting.
Archives and backupsFormat, date, custodian, encryption, restore method and destruction date.Operational backup, legal archive or redundant copy to remove.

For every data set, record the business owner, technical custodian, data type, sensitivity, source authority, update cadence, users, contractual restrictions and intended destination. A folder named “final contacts” is not a reliable source-of-truth designation.

The NIST Cybersecurity Framework 2.0 describes high-level cybersecurity outcomes rather than prescribing one implementation. Its asset-management and data-security outcomes support maintaining inventories, managing information through its lifecycle, protecting confidentiality, integrity and availability, and testing backups. Use those outcomes as questions for the migration—not as a claim that a particular CRM is compliant or secure.

3. Decide what moves, what is archived and what is removed

Classify each field and record set into one of four dispositions:

  1. Migrate: needed in the target CRM for an approved operational, reporting or retention purpose.
  2. Transform: needed, but must be standardized, split, combined, relabeled or converted before import.
  3. Archive: retained outside the active CRM for a documented reason, with an owner, access rule and destruction date.
  4. Dispose: redundant, stale or unsupported by a current business or legal need, subject to approved retention and disposal rules.

The FTC's business guide to protecting personal information recommends knowing what personal information a business holds, keeping only what it needs, limiting access and disposing of information securely when it is no longer needed. The guide also recommends a written retention policy when information must be kept. These are security principles, not a universal retention schedule.

Do not let a software provider decide retention obligations. Requirements can differ by industry, contract, record type and legal context. The project owner should obtain qualified legal or compliance guidance where needed, then translate the decision into a field-level rule the migration team can execute.

Document exclusions. If old email bodies, unstructured notes, attachments or lost opportunities will not move, record why, where any approved archive lives, who may access it and how staff will respond when historical information is needed.

4. Build a field-mapping workbook

The field map is the core control document. Give every source field one explicit destination or disposition. Avoid “the vendor will figure it out during import.”

Recommended columns for the migration workbook
ColumnPurpose
Source system, object and fieldIdentifies the exact origin, not just a friendly label.
Source definition and ownerExplains what the value means and who can resolve ambiguity.
Target object and fieldNames the exact destination and target definition.
DispositionMigrate, transform, archive or dispose.
Transformation ruleDefines formatting, normalization, lookup, split, merge or default behavior.
Allowed valuesLists picklist values, date format, currency, units and null behavior.
Sensitivity and accessRecords classification and which roles should see or edit the value.
Source of truthResolves conflicts when the same value appears in multiple systems.
ValidationStates the count, query, sample or business rule that proves success.
Exception ownerNames who decides when a record does not fit the rule.

Map identifiers before names. Customer, organization, opportunity, project and invoice records may use different keys across systems. Define a crosswalk that preserves the original identifier, creates the target identifier and allows the team to trace a migrated record back to its source without relying on a name match.

Write null and default rules explicitly. A blank value is different from “no,” “unknown,” “not applicable” and a system-generated default. Filling blanks with a convenient value can make a report look complete while changing its meaning.

5. Set data-quality and duplicate rules before cleanup

Do not begin with an unrestricted merge tool. Define how the business recognizes the same person or organization and which source wins when values conflict.

  • Exact duplicates: same approved unique identifier.
  • Probable duplicates: combinations such as normalized email, phone, name and organization that need review.
  • Related but distinct records: household members, multiple locations, shared inboxes, parent and subsidiary organizations or repeat engagements.
  • Conflicting records: different owners, consent states, lifecycle stages or contact details that require an accountable decision.

Choose survivorship rules field by field. The newest timestamp may be appropriate for a phone number but not for a consent record, contract term or closed-work status. Preserve source, timestamp and reviewer evidence for material changes.

Create measurable quality thresholds without inventing perfection. Useful checks include required-field completeness, valid formats, allowed-value conformance, orphaned child records, duplicate-review backlog and records that fail the transformation rule. Report exceptions separately; do not hide them inside a percentage.

6. Control vendor access and data transfers

A migration often creates extra copies of sensitive information: source exports, transformation files, test imports, error logs, screenshots and backups. Treat each copy as governed data with a purpose, owner, access rule and deletion date.

The FTC's current vendor guidance recommends assessing supplier risk, writing security expectations into contracts, limiting access to what a vendor needs, using multifactor authentication and encryption, verifying compliance and addressing retention and deletion. The New York Attorney General's SHIELD Act overview states that businesses maintaining New York private information must use reasonable administrative, technical and physical safeguards, including selecting capable service providers and requiring safeguards by contract.

For the migration, record:

  • which people and service accounts can export, transform, import or delete data;
  • whether access is read-only, temporary and protected by multifactor authentication;
  • where files are stored, encrypted and transmitted;
  • whether test environments contain production data and how exposure is minimized;
  • which subcontractors, connectors or processing locations are involved;
  • what activity and administrative logs will be retained;
  • how incidents must be reported and investigated; and
  • when vendor access, exports, test data and temporary credentials will be removed.

Use masked or synthetic data for early configuration when practical. When production records are required for a migration test, use the smallest approved sample that can exercise the necessary relationships and edge cases. A shared spreadsheet attached to email should not become the default migration workspace merely because it is convenient.

This section is a planning framework, not a legal or security determination. A business handling regulated or contract-restricted information should have the relevant controls reviewed by qualified professionals.

7. Run staged test migrations

Use at least three stages: a structural sample, a representative sample and a full rehearsal. Reuse the same approved mapping rules and make changes through version control or a dated change log.

Structural sample

Import a small set covering each object, relationship and field type. Confirm that identifiers, owners, dates, notes, attachments, picklists and parent-child links behave as designed. The objective is to expose mapping and configuration errors quickly, not to demonstrate scale.

Representative sample

Select cases that reflect real variation: active and inactive customers, multiple services, missing values, duplicates, long notes, international formats where relevant, failed inquiries, closed work and records with restricted access. Include deliberately difficult cases and rejected records.

Full rehearsal

Run the complete extraction, transformation and import in a non-production target or isolated rehearsal environment where the project allows it. Measure duration, error volume, reconciliation results, integration behavior and staff validation time. Rehearse the cleanup and reset procedure as well as the import.

Evidence to capture for every migration test
Test areaEvidence
CountsSource, excluded, transformed, imported, rejected and reconciled records by object.
RelationshipsSampled parent-child, contact-organization, opportunity-customer and activity links.
ValuesRequired fields, allowed values, dates, currency, multiline text and null behavior.
Ownership and accessRecord owners, teams, role visibility, edit rights and denied-access cases.
WorkflowRouting, tasks, notifications, integrations and reports using migrated records.
ExceptionsRejected records, reason, correction, approver and rerun result.
PerformanceObserved import and validation time under the tested conditions, without generalizing beyond them.

Keep the source extracts unchanged. Perform transformations on controlled working copies so the team can reproduce a result and distinguish source problems from transformation errors.

8. Plan cutover, the change freeze and rollback

A cutover plan should identify the last safe point for source-system changes, the final export time, integration pause, import sequence, validation window, go-or-no-go owner and user communication. Because this article does not authorize customer messages, communication planning means preparing internal responsibilities and approved templates—not sending anything.

Define how new activity will be handled during the freeze. Options may include a short controlled pause, a documented catch-up import or dual capture with reconciliation. Avoid uncontrolled dual entry: when staff can edit both systems without a reconciliation owner, neither remains authoritative.

A rollback plan is more than keeping the old login active. Record:

  • the failure conditions that trigger rollback;
  • who can make the decision and by what deadline;
  • how target-system changes will be isolated or exported;
  • how integrations and user access return to the prior state;
  • how work completed during the cutover window is reconciled; and
  • which backups and restore steps have been tested.

The NIST CSF 2.0 Small Business Quick-Start Guide is designed to help smaller organizations begin cybersecurity risk management. It does not prescribe a CRM cutover. Orbital's rollback and staging steps are practical project controls aligned with the broader need to protect operations and recover from disruptions.

9. Validate the business outcome, not only record totals

Counts matter, but identical totals do not prove that the data is usable. Acceptance should combine automated checks, sampled records and real workflow tests performed by accountable business users.

Migration acceptance matrix
DimensionExample acceptance evidence
CompletenessApproved source population reconciles to migrated, archived, disposed and rejected totals.
AccuracyRisk-based samples match source values and approved transformations.
RelationshipsContacts, organizations, opportunities, activities and files remain connected as intended.
PermissionsNamed roles can see and change only the approved records and fields.
WorkflowA new inquiry, follow-up, handoff, status change and completed-work path function end to end.
IntegrationsApproved systems send, receive, retry and report errors according to the documented source-of-truth rules.
ReportingOwners can explain key pipeline definitions, exclusions and differences from the legacy report.
ExceptionsEvery material exception has an owner, disposition and due date.
RecoveryBackup, restore or rollback evidence meets the project's approved test scope.

Have sales, service, finance or operations representatives validate the records they actually depend on. The project team should not sign off on behalf of a business process it does not own.

Compare reports carefully. A difference can reflect a migration defect, a corrected duplicate, a new stage definition, an excluded historical period or a legacy calculation that was never documented. Reconcile the definition before deciding which number is “right.”

10. Require a migration evidence pack and vendor exit

Completion should include a dated package that another qualified administrator can understand:

  • approved decision brief and scope;
  • source-of-truth and system inventory;
  • final data disposition and retention decisions;
  • field map, transformation rules and version history;
  • identifier crosswalk and duplicate policy;
  • test plans, results, exception log and approvals;
  • final reconciliation counts by object;
  • cutover timeline, go-live decision and rollback plan;
  • access, service-account and integration register;
  • backup, archive and restore information appropriate to the project;
  • known limitations and post-launch backlog;
  • vendor deletion and access-removal confirmation; and
  • owners for data quality, permissions, workflows and future changes.

Do not accept a vendor-owned export account or an undocumented import script as the only path to the data. Confirm what the business owns, which reusable materials it receives, what proprietary dependencies remain and how a future export will work.

The NIST Privacy Framework getting-started guidance encourages organizations to consider privacy risk across the data lifecycle from collection through disposal. Use that lifecycle perspective to close the temporary migration environment rather than leaving exports, test users and elevated access behind after launch.

11. Use a 30-day stabilization plan

Days 1–3: protect operations

  • monitor imports, integrations, automation errors and user access;
  • reconcile new inquiries and other time-sensitive records;
  • route defects through one triage owner; and
  • preserve evidence before changing mapping or transformation rules.

Days 4–14: validate real use

  • review duplicates, rejected records and missing relationships;
  • sample completed workflows across responsible teams;
  • compare reports with the approved definitions;
  • remove temporary administrator and vendor access that is no longer required; and
  • train staff on the new source of truth, escalation and correction process.

Days 15–30: close the migration

  • resolve or accept remaining exceptions with named owners;
  • confirm archive, retention and disposal actions;
  • obtain approved deletion evidence for temporary vendor copies where applicable;
  • record baseline data-quality measures and future review cadence; and
  • separate migration closure from the next enhancement backlog.

Use Orbital's business software integration checklist when the remaining risk is in CRM connections, and the tech stack audit checklist when ownership, subscriptions or duplicate tools remain unclear.

Orbital's business take

A successful CRM migration is not the largest possible copy of the old system. It is a traceable transfer into an approved operating model, with enough evidence for the business to trust its records, permissions, workflows and reports—and a recoverable path when something fails.

Explore Orbital's tech stack consulting, connected-business planning, automation and technology strategy, sales and CRM systems consulting and owner oversight planning. A 60-minute consultation can help turn an existing migration idea into a decision brief and risk register; software, implementation, data work, integrations, legal review and ongoing administration are separately scoped.

Sources and editorial limits

Primary sources were checked November 4, 2026. The FTC supplies current small-business cybersecurity and personal-information guidance. NIST supplies the voluntary Cybersecurity Framework 2.0, its Small Business Quick-Start Guide and Privacy Framework resources. The New York Attorney General supplies the current SHIELD Act overview. The migration brief, source register, disposition categories, field-map columns, duplicate policy, staged tests, acceptance matrix, cutover plan, evidence pack and 30-day stabilization plan are Orbital editorial recommendations—not requirements from those organizations. Guidance, laws, platform capabilities and contracts can change. Orbital did not access systems, inspect or migrate customer records, test a CRM, determine legal or regulatory obligations, audit security, measure search volume, estimate cost or time, or guarantee data quality, adoption, security, compliance, revenue or any business outcome.

Primary source links

Your next chapter.
Let’s give it lift.

Start a conversation