Setup and migration

Switch without losing appointments.

Check your records, settings, and future bookings before the move. Rehearse the handoff before going live.

First checks

  1. Model the business first

    Define the service, people, places, time rules, policies, and customer journey before touching import tools.

  2. Move meaning, not rows

    Preserve identities, relationships, statuses, time zones, money references, forms, history, and known omissions.

  3. Rehearse away from live data

    Use invented records to prove import, staff work, customer booking, cancellation, export, and recovery.

  4. Keep the old route recoverable

    A cutover needs a freeze, reconciliation, stop trigger, rollback path, owner, and controlled reopening.

Two adult operators map an old paper calendar and record cards into a configured system, test fictional records, compare both views with a checklist, and keep the previous process available during cutover.
Map, test, compare, and keep a way back.A safe migration proves services, people, customer records, future bookings, and changes with invented test data before the live process moves.
Synthetic setup-readiness checker

What must happen before a cutover is scheduled?

Choose the current evidence state for the product-neutral plan. The result names the next checkpoint; it does not authorize a live migration.

Your synthetic setup result

Run the reversible rehearsal next

A complete mapping has not yet passed a reversible synthetic import rehearsal.

Requirements
complete
Configuration
complete
Source inventory
complete
Field mapping
complete
Import rehearsal
not-run
Staff-role test
not-run
Customer-journey test
not-run
Recovery proof
documented-only
Training
partial
Cutover owner
unnamed

Next: Run the made-up booking example away from live customer records and reconcile every expected count and relationship.

The checker stores nothing and cannot inspect a product, export, account, backup, role, customer journey, or rollback. A ‘ready to schedule’ result still requires current product evidence and a human go or no-go decision.

Why onboarding checklists are not enough

Configuration, migration and launch are three different jobs.

Configuration defines future behavior. Migration carries selected old records into that model. Launch changes which route customers and staff rely on. Combining them hides failures until the business is live.

The cutover rule

Safe cutover = complete target model + preserved source inventory + explicit mapping + reversible rehearsal + role and customer tests + reconciled recovery + trained owner

A synthetic rehearsal can contain 12 clients, eight future appointments, two cancelled appointments, three staff roles, two services and one location. Counts alone do not pass: identities, relationships, times, statuses, permissions, messages, exports and rollback must also agree.

Follow all eight controlled stages

A booking page is not ready when the logo looks right. Services, staff, locations, policies, records, permissions, messages, calendars, payments, training, recovery, and the old system must agree before customers are moved.

Build a reversible setup and cutover plan without naming a winning product.

  1. Freeze the current operating truth

    Record how customers book today, who changes appointments, where time and money are authoritative, which manual exceptions exist, and what cannot stop during a transition.

  2. Build the destination skeleton

    Create a product-neutral manifest for business identity, services, duration, price references, staff, locations, resources, availability, policies, forms, notifications, calendars, payments, permissions, and reports.

  3. Inventory the source records

    Preserve an untouched export and count clients, future and past appointments, cancellations, recurring items, packages, forms, notes, payment references, messages, audit history, users, and configuration.

  4. Map fields, identities and omissions

    For every required object, name the source field, destination field, transformation, stable identity, relationship, status meaning, time-zone rule, omission, exception, and reviewer.

  5. Run a reversible rehearsal

    Import the fixed invented fixture into an isolated destination, inspect rejected and duplicate rows, reconcile every expected record, then undo or restore the rehearsal.

  6. Test staff and customer journeys

    Use separate invented roles to create, change, cancel, notify, sync, collect no real money, report, export, deny unauthorized actions, and recover failures.

  7. Train the operating owners

    Give each role the tasks it performs, actions it cannot perform, escalation route, recovery queue, cutover responsibilities, and old-system fallback.

  8. Schedule a controlled cutover

    Take a fresh snapshot, freeze writes, import only the approved delta, reconcile, communicate, monitor stop triggers, keep rollback available, and reopen only after a human go decision.

Carry one approved workflow from setup into stable operation.

Use the same scope, owners, evidence, stop conditions, and recovery path from configuration through handover. Each phase must finish with a reviewable output before the next phase begins.

01

Freeze the operating truth

Business process owner

Record the current booking journey, exceptions, authorities, records, integrations, customer fallback, and work that cannot stop.

Exit: Approved current-state map, implementation scope, protected source inventory, known unknowns, and one accountable change owner.

Stop: A required workflow, exception, authoritative record, owner, or continuity need is still unknown.

02

Configure the destination model

Configuration owner

Translate approved services, staff, locations, resources, time, money, forms, permissions, messages, calendars, and reports into a product-neutral manifest.

Exit: Versioned configuration manifest with named specialist handoffs, acceptance checks, omissions, and an isolated test destination.

Stop: Configuration is being improvised inside an import, product demo, or live customer route.

03

Rehearse migration and recovery

Migration and recovery owner

Map identities and relationships, use invented records, preserve rejects and duplicates, reconcile the expected result, then reverse or restore it.

Exit: Saved mapping, fixed fixture result, exception ledger, reconciled export, proved recovery path, and usable old-route fallback.

Stop: The team cannot explain a mismatch, rerun safely, restore the test, or return to the previous route.

04

Train roles and stage the launch

Training and access owner

Give each role its normal tasks, denied actions, escalation path, recovery queue, customer fallback, and cutover responsibility.

Exit: Role-specific task rehearsal, tested denials, training record, support handoff, customer communication rule, and named fallback coverage.

Stop: Training is only a vendor tour, administrator access is shared, or the team cannot perform and recover its real tasks.

05

Cut over—or stop

Cutover decision owner

Control the final snapshot, write freeze, approved delta, public route, communication, reconciliation, stop trigger, rollback, and human go or no-go decision.

Exit: Signed checkpoint record with the fresh source state, reconciled target, current fallback, explicit residual risk, and final decision.

Stop: Any must-have journey fails, records disagree, recovery is unavailable, ownership is absent, or a stop trigger is reached.

06

Stabilize and hand over

Continuing operational owner

Reconcile bookings, money, messages, calendars, automation, permissions, reports, exports, support work, and the controlled old-system closure decision.

Exit: Accepted operating runbook, open-issue queue, monitoring and recovery ownership, dated export, change process, and separately approved source-retirement decision.

Stop: A clean dashboard is being treated as proof while customer, staff, financial, integration, or exit records remain unresolved.

This blank playbook prepares a product-neutral implementation record. It does not configure an account, move customer data, prove a product, train staff, authorize a launch, perform a rollback, approve source deletion, or replace privacy, security, legal, financial, accessibility, or qualified operational review.

Review all setup domains

Business identity and booking page

Name, contact, location display, brand, language, time zone, public routes, search visibility, and fallback contact.

Specialist handoff
Booking-page UX and website embedding own the rendered customer path.

Services and policies

Service identity, duration, price reference, eligibility, lead time, horizon, cancellation, rescheduling, deposits, and exceptions.

Specialist handoff
Use the deposits and balances, cancellations and rescheduling, recurring appointments, packages and memberships, classes and waitlists, staff availability, resource capacity, service areas, and buffers and travel guides for the detailed money, capacity, recurrence, and time rules.

Staff, locations and resources

Distinct identities, assignments, working hours, time off, locations, rooms, equipment, travel, and conflict rules.

Specialist handoff
Use the staff availability, resource capacity, service areas, and buffers and travel guides for detailed availability calculations; this page checks that the chosen model is configured and reviewed.

Roles and administration

Owner, manager, receptionist, provider, support, exporter, high-risk actions, denials, reviews, and offboarding.

Specialist handoff
The staff-permissions guide owns the detailed least-privilege matrix.

Customer records and forms

Client identity, contact fields, consent or policy references, custom questions, notes, edits, history, retention, and deletion boundaries.

Specialist handoff
Form fields own answer integrity; export owns portable scope and completeness.

Messages, calendars and payments

Trigger, destination, templates, calendar direction, payment boundary, retries, failures, and final reconciliation.

Specialist handoff
Use the reminders and confirmations, calendar synchronization, payment integration, and checkout reconciliation guides for detailed delivery and money rules; this page verifies the launch handoffs.

Migration and reconciliation

Source snapshot, field map, import batch, rejected rows, duplicates, counts, identities, relationships, time zones, money references, and rollback.

Specialist handoff
The export guide owns data completeness; this page owns the cutover sequence.

Training, maintenance and exit

Role training, runbooks, monitoring, support, change review, backups, recovery, export refresh, contract exit, and old-route closure.

Specialist handoff
Reliability and build/buy guides own continuing operational responsibility.
Reconcile the complete migration matrix
Record classPreserveValidateRecover
ClientsStable client identity, contact fields, notes, duplicate rulesCounts plus sampled field and duplicate reconciliationRestore destination or undo the batch; keep the source untouched
Future appointmentsExact instant, source zone, service, staff, location, status, customer and policyEvery upcoming item appears once in customer and staff viewsReturn customer traffic and staff edits to the frozen source
Past and cancelled appointmentsHistory, cancellation meaning, audit references and reporting needRequired history is either imported or explicitly retained elsewherePreserve the source archive and documented lookup route
Recurring, classes and creditsSeries or occurrence identity, attendee, entitlement, balance and exception meaningRelationships and remaining value reconcile, not only row totalsStop the migration when the destination cannot represent required meaning
Forms, notes and attachmentsField identity, answer meaning, permissions, retention and unavailable formatsApproved samples remain attributable and readable to the right rolesRetain the protected source or choose an approved manual handoff
Payments and financial referencesBooking obligation, transaction and refund references—not card secretsBooking, receipt, gateway and export references reconcileDo not move or recreate a payment when state is uncertain
Users, roles and integrationsSeparate identities, permissions, calendars, message routes, webhooks and ownershipAllowed and denied actions plus disconnect and recovery paths passRevoke the new access and restore the old authorized route
Configuration and public linksAvailability, policies, templates, URLs, embeds, redirects and search stateCustomer, staff and fallback journeys reach one current resultSwitch links back and preserve the last known working configuration
Handle changes and cutover red flags
01

A service or policy changes during preparation

Version the destination model and map the delta

Do not import against a stale service, duration, price, cancellation, payment, capacity, or form definition.

02

A new booking arrives after the source snapshot

Keep a controlled delta ledger

Name the freeze time, late-write owner, delta import, duplicate key, and final reconciliation instead of copying records ad hoc.

03

A staff member or location changes

Recheck assignment and permissions

Availability, resource, calendar, message, reporting, role, training, and offboarding implications may all change.

04

An import rejects or duplicates records

Stop the batch and preserve exceptions

Do not hide rejected rows, edit live records until counts match, or rerun without a stable idempotency and cleanup rule.

05

The destination changes its plan or behavior

Recheck the exact current route

Plan limits, fields, imports, roles, messages, integrations, exports, and cancellation terms require a new dated source and rehearsal.

06

The old system approaches closure

Prove the exit before deleting access

Keep required history, exports, configuration, contracts, invoices, audit evidence, deletion decisions, and support contacts under an accepted owner.

The launch date exists before the source inventory

The team has promised a switch without knowing which bookings, clients, statuses, relationships, histories, or exceptions must survive.

Row counts are the only reconciliation

Matching totals can still hide wrong customers, duplicate appointments, shifted time zones, missing relationships, changed statuses, or broken permissions.

The import is first tried on live customer data

A failed mapping, duplicate run, unexpected message, payment action, or calendar write can create a real customer and staff incident.

Everyone receives administrator access

Shared or excessive authority hides attribution, weakens denials and offboarding, and can expose customer or configuration data unnecessarily.

The old route is closed before recovery is proved

A written rollback cannot restore service when exports, configuration, credentials, contracts, DNS, embeds, integrations, or support access are incomplete.

Training means watching the vendor tour

Staff have not demonstrated their real create, change, cancel, message, calendar, payment-boundary, report, denial, escalation, and recovery tasks.

Prove the handoffs before touching a live booking.

Use the same invented records and expected outcomes from source inventory through destination setup, import, role tests, customer tests, export, reversal, and reconciliation.

  1. Freeze a synthetic source with 12 clients, eight future appointments, two cancelled appointments, three staff roles, two services, one location, one form, one calendar handoff, and no real payment or message delivery.
  2. Configure the destination skeleton, map every required identity and relationship, import once, preserve rejects and duplicates, and reconcile counts, samples, time meanings, permissions, exports, and notifications without using real people.
  3. Reverse or restore the rehearsal, prove the old route remains usable, record training and cutover ownership, then keep the live go or no-go decision closed until a human reviews the complete evidence.

Next steps