Leaving Workday without losing the record

Leaving Workday is an enterprise extraction, and the honest first step is to decide whether you should. Most organisations running Workday are large enough that their next system is another enterprise suite, because they need the scale, the finance integration and the configurability that defines this category. A move to a lean, record-first product is the right call only for an organisation that has concluded the enterprise suite is heavier than it now needs, a divested or simplified business, or a smaller entity being carved out of a group. If that is you, this guide is for you. If you still need enterprise-scale HCM and finance in one platform, a lighter product is the wrong destination, and the most useful thing we can tell you is to stay in the category.

One note on sourcing. What follows about Workday’s own product is our reading of their public material, read in September 2026, and gathered through search because direct access to their documentation is blocked from where this was written. Their documentation is the authority, and you should confirm the current article before planning around a feature or a report. Everything else here is about your obligations and your data, which do not move when a vendor ships a release.

Start by agreeing who owns the configuration

An enterprise deployment is configured, and usually run, with implementation and support partners, so the knowledge of how your tenant works may sit outside your own team. Before anything, write down who holds which obligation: who owns the tenant, who built the configuration, who runs the payroll cycle, and who can generate the custom reports and the extracts you will need. Agree in writing that the configuration documentation and the report definitions are handed over while the relationship is warm. After notice, you are negotiating for the knowledge with the least leverage you will have.

What you are actually moving

• The worker master, with every custom field, and the organisation, position and cost-centre structures that give it meaning.
• Compensation history, grades, plans and the record of changes.
• Payroll output, registers and payslips for every closed period, per country.
• Statutory output, the filings you made, their references, and the year-end documents, per country.
• Performance and talent records, reviews, goals, calibration history, as far back as you want to carry.
• Leave balances and policies, the numbers and the accrual rules.
• Documents, everything held against a worker.
• The configuration and the custom reports, captured as artefacts, because they define how the system worked and do not export as a single file.
• The year-to-date position, per person, the running totals that carry through each country tax year.

The export checklist, in order

Do all of this while the tenant is live and access is intact.

1. The worker master and the structures, full detail with every custom field, as your index.
2. Reports as data. An enterprise suite exposes most of what you need as reports; export each as data, not as a formatted document, so it loads into whatever comes next. Pull the standard extracts and the custom reports you rely on.
3. Payroll registers and payslips, per closed period, per country, filed by period.
4. The statutory set, per country and period, each filing paired with its reference, plus year-end documents.
5. Compensation, performance and talent history, exported as far back as you want to carry.
6. Leave balances and policies, numbers exported, rules written out in plain words.
7. Documents, tested on one worker before you size the job across headcount.
8. The configuration, captured in writing, the business processes, security groups, custom fields and report definitions, in plain language with screenshots where a path is complex. This is the specification for rebuilding, and nobody can generate it after the tenant closes.
9. The year-to-date pack, one spreadsheet per country, one row per person, reconciled before you accept it.

Timing: the tax year and the fiscal calendar decide

For each country, the cheapest payroll boundary is the start of that country’s tax year, with everyone at zero year to date and the prior year closed by the system that computed it. If finance data is entangled with HR in the same tenant, the fiscal calendar matters too, so coordinate with finance before you set a date. Avoid every year-end and period-close run-in, when reconciliation and statutory deadlines compete with the migration and the deadlines win. A multi-country enterprise will not have one neat boundary, so sequence the payroll cutover country by country and keep the record move separate.

The separation that makes a staged move possible

The system that holds your records and the system that computes your payroll do not have to be the same system, and on a multi-country move that separation lets you stage the work rather than attempting one cutover everywhere at once. Move the record when it suits you; change who computes at each country’s boundary.

That separation is Capstan’s default. Capstan compiles joiners and leavers, pay structures, leave, attendance and adjustments into a documented export for your payroll partner, and files the results the partner returns back onto each employee record. It holds no rate, slab, bracket, threshold or formula in any country, computes no gross to net, and files nothing with any authority. The payroll module sets out what crosses the line in each direction.

Where it tends to be hard

The configuration is the project, not the data. The records export; the business processes, security model and custom reports do not. Budget most of your time for rebuilding, and be realistic that a lean product will not reproduce enterprise configurability, which is usually the point of moving.

History is deep and will flatten on export. Pull the change history and the as-of snapshots you need as deliberate, separate extracts; after the tenant closes, they are gone.

Finance entanglement. If HR and finance share the tenant, agree with finance what is an HR record and what is theirs before you extract anything.

Access ends on a schedule. Read your contract for the retrieval window and the deletion process, diarise both, and finish well before the first.

What you have to keep, whoever you buy from

Employment and tax records carry retention obligations that follow the employer, not the software, and they differ by country and statute. Keep your own dated archive outside any vendor, per country and tax year, sub-foldered by period, holding the registers, payslips, filings and year-end documents. Confirm the periods with your advisers rather than with any vendor, including us.

Standing up the replacement

If a lean product is genuinely the right destination, use the implementation checklist for the order and read this as the overlay. Validate the export first, then load in parallel while the old tenant is live. With Capstan that is a spreadsheet import with a dry run that previews every row and fails bad rows individually; documents load per person, opening leave balances are entered as adjustments with a reason, and effective-dated history is not imported, so your export stays your archive of it. What imports and what does not is written out in full. Set one go-live date per entity, hand your payroll partner a past period of inputs per country and compare, and reconcile ten people by hand across structures before you cancel.

Who should stay where they are

Be fair to the product you are leaving, and to yourself. An enterprise platform spanning HR, finance, planning and analytics at scale is doing a large job that a lean product does not attempt, and if you still need that scale, staying is the right answer and we would say so plainly. A focused, record-first system with partner-integration payroll is the right destination only when the enterprise suite has become heavier than the organisation needs: a carved-out entity, a simplified business, a team that has decided it wants a portable record and a price it can total itself rather than an enterprise contract. If you are not in one of those situations, moving to a lighter product is a downgrade, not a simplification. The Workday comparison sets the two published models side by side, including where Capstan is the wrong tool, and HRIS, HRMS and HCM explains where each sits.

Where to go next

If you run across entities, multi-entity HR covers holding one record across them. Whatever you buy, start your shortlist with the export is the habit worth keeping, and read how to read an HR pricing page before you compare totals. Capstan’s figures are on one public pricing page.