Setting up a new payroll customer
- Create the Tenant and its Payroll (Organization Unit).
- Assign Regulations to the Payroll - country, chart of accounts and accounting document included.
Important: if you're planning test runs for past periods, set
accountingStartDateto the start of those test runs, not to the actual go-live date. - Enter company master data, case by case.
- Enter Cost Centers, if the company uses them.
- Activate or copy any special or missing wage types.
- Complete WageType-Accounting-Allocation (and the chart of accounts, if needed).
- Migrate and import employee data as of a cutoff date.
- If no
employeeIdentifierexists yet, it needs to be captured automatically. - Guidance on how the data must be provided (e.g. via Abaconnect, Excel, or normalized data feeds).
- To let platforms do their own field matching, Ason provides a field list - changes to it are always visible in the release notes.
- Missing data should be retrievable directly via the API, at field level. To avoid ending up with a flood of one-off cases/fields, there should be standard employee templates from the start.
- If no
- Optional: run test or parallel runs, comparing results against the legacy system.
- Import mutation data for every employee, per test period.
- Compare Ason's cumulative journal against the legacy system's.
- If those test runs used incomplete data, rebase
accountingStartDate- see Accounting Start Date on the Payroll page for what that does and its limits. - Go live - daily business with Ason begins.
Testing before go-live
- Steps 8 and 9 exist so you can run test periods (e.g. November/December of the previous year) even though results won't be fully correct since earlier months are missing
- What matters: current company and employee data is already in the system, so the setup gets exercised end-to-end before the real go-live
- Once satisfied, rebase
accountingStartDateto wipe those provisional periods before starting for real
Migration tooling
Ason is also working on a dedicated migration tool to make this process more self-service, rather than the current mostly-manual, API-driven flow. More details on this soon.
