MIGRATION WORKBENCH
Clean files. Clear mappings. Fewer surprises.
A field service CSV import starts before the upload. Use this staging schema to understand your records, preserve relationships and make exceptions visible.
A staging workbook, not a magic import format
Create three tabs: customers, services and upcoming jobs. Keep the original exports untouched and work on a dated copy. The field names below are a planning convention, not a guarantee that CrewViper or another importer accepts these exact headers. Map each column against the actual import preview and retain unsupported information separately.
| Tab / field | Example | Check before mapping |
|---|---|---|
| Customers: source_customer_id | C-1042 | Required staging key. Preserve as text; never generate a new key just because the display name changed. |
| Customers: display_name | Morgan Sample | Use a fictional record for testing. Do not combine two households because names match. |
| Customers: email / phone | Separate contact columns | Keep phone numbers as text. Confirm which contact is primary and retain additional contacts separately if unsupported. |
| Customers: service_address | Separate street, city, region, postal code | Preserve postal codes as text. Multiple properties need explicit mapping, not an overwritten address. |
| Services: source_service_id / name | S-020 / Garage reset | Use stable identifiers; names alone may collide. |
| Services: unit_price / currency | 125.00 / USD | Numeric amount without a currency symbol. Blank is unknown, not a zero-dollar service. |
| Jobs: source_job_id / source_customer_id | J-110 / C-1042 | Every linked customer key must resolve exactly once. |
| Jobs: planned_date / start_time / timezone | 2026-12-08 / 09:00 / America/Denver | Keep the business timezone explicit. A date-only job is not a confirmed appointment time. |
| Jobs: source_status / assigned_crew | Scheduled / Crew A | Map statuses deliberately. Confirm the destination crew exists and has appropriate access. |
One worked relationship check
A staging job J-110 references customer C-1042 and service S-020. Before importing that job, verify that its customer mapping points to one destination customer, its service price has been reviewed, and Crew A is a valid destination assignment. If C-1042 occurs twice, stop that row for review rather than choosing whichever appears first.
Maintain a separate mapping ledger: source type, source ID, destination ID, batch ID, outcome and reviewer. Outcomes should distinguish created, matched, skipped and rejected. A skipped duplicate is not automatically a successful match.
Run a small batch before the full file
- 01
Inventory the source
Record file names, export dates, row counts and business timezone. Identify open quotes, unpaid balances and recurring jobs that require separate handling.
- 02
Preview representative rows
Include a duplicate name, a leading-zero postal code, an empty price, multiple properties and a rescheduled job. Inspect mapping warnings before committing.
- 03
Reconcile the result
For each source row, record an outcome. Check customer links, totals, dates and assignments in the destination, not only the success count.
- 04
Prove safe retries
Confirm what happens if a batch is retried. Do not assume the importer updates existing records or prevents duplicates without a controlled test.
Protect the files while you work
Customer exports contain personal information. Restrict access to the migration team, use approved storage, and keep real records out of screenshots and public demos. Treat spreadsheet cells beginning with formula characters as untrusted content; inspect source data without executing embedded formulas. Agree on retention before making additional copies.
Do not infer a payment from an invoice export. Do not convert a missing phone number into a dummy number. Preserve unknowns and resolve them with the owner.
Next: use the cutover and rollback checklist to decide when the cleaned files are ready for production.
Test the workflow before moving the business.
CrewViper is in private pilot. The public demo uses fictional data and resets on reload. It is not an importer, a backup test, or proof that a production integration is connected. Do not upload real customer data to the demo.
