Skip to content

ATS Data Migration: Move Your Recruiting Database Without Losing Records

Switching recruitment software raises one practical question before anything else: can the candidates, clients, notes and resumes you have built up over years survive the move intact?

This guide explains the two routes available for bringing an existing recruiting database into a new ATS, what each route requires from you, and how to check the result before you rely on it.

Two routes into a new ATS

> Every migration decision starts with a simple split: do the import yourself with a built-in tool, or hand the work to a migration team. The right choice depends less on record count than on how messy your source data is and how much internal time you can spare.

Self-service import. A built-in Excel/CSV importer handles candidates, companies/clients and the notes attached to them. Resume files can be uploaded separately and parsed into structured fields by an AI resume parser. This route costs nothing beyond your own time and suits teams with reasonably tidy exports.

Managed migration. A dedicated team of data migration engineers takes the full dataset and loads it for you. Pricing scales with the volume of data to be imported, and the work typically runs between 10 and 12 business days. This route suits teams with large, mixed or poorly structured exports, or teams that cannot dedicate staff hours to mapping spreadsheets.

Neither route is inherently better. The self-service path gives you control and immediacy; the managed path trades money for hands-off handling of complexity.

How a managed migration runs, step by step

> A managed migration follows three stages. Understanding them helps you plan downtime and assign an internal owner.

  1. Data discovery. Your existing ATS data is extracted into a spreadsheet or external file. This is the stage where format problems surface, so it is worth confirming early that your current provider can produce a complete export rather than a partial one.
  1. Importing. The receiving provider performs the initial load of the records you supplied. Field mapping decisions happen here: which source column becomes which destination field, and what happens to records that do not fit the target structure.
  1. Deployment and review. After export and import, an engineer runs a quality check. Once that passes, the new system is ready for use.

A useful habit is to nominate one person on your side who owns the export, signs off the field mapping and runs the acceptance check at the end. Migrations stall most often when nobody internally can answer questions about what a given column was supposed to mean.

Accepted formats and how much data survives

> The service accepts any size of data in CSV, XLS, XLSX, .BAK, .BACPAC, .SQL and database formats. The stated expectation is that 99% of your data will be cleaned and transferred.

That 99% figure deserves a moment of thought rather than blind trust. In practice, the records most likely to need attention are the ones with inconsistent structure: free-text notes crammed into a single field, duplicate candidate entries, or custom fields with no obvious destination. Planning for a small reconciliation pass after import is realistic; expecting a byte-for-byte replica is not.

If your export is a database backup (.BAK, .BACPAC, .SQL) rather than a flat file, confirm with the migration team which tables they need before you send anything, since a full backup may contain far more than the recruiting records themselves.

Security, ownership and compliance during the move

> Two concerns come up in almost every migration conversation: is the data safe in transit and at rest, and who controls it afterwards. The statements below describe the source platform's published position; confirm each against your own provider's current documentation before you commit.

Storage and encryption. The source platform states that data it holds is stored in data centres managed by Amazon Web Services and encrypted with AES-256 bit encryption, described there as the global industry standard for internet data security, with regular backups taken.

GDPR. The source platform states that it is GDPR compliant and provides tools intended to help your own organisation meet its GDPR obligations.

Ownership. The source platform states that you own all the data you put in, that its role is to help you manage it, and that you can extract your data at any time.

For a migration specifically, ask two extra questions that general statements do not answer: how the export file travels between systems, and how long any intermediate copy is retained. Those are operational details worth confirming with the migration team before you hand over a full database.

Pre-migration checklist

> Work through this before the export begins.

  • [ ] Confirm your current provider can produce a complete export, not a filtered or truncated one.
  • [ ] Decide which record types matter: candidates, companies/clients, notes, resume files, and any custom fields.
  • [ ] Identify duplicate records in the source and decide whether to merge them before or after import.
  • [ ] List your custom fields and note the intended destination field for each.
  • [ ] Choose the route: self-service import or managed migration.
  • [ ] Nominate one internal owner for export, mapping sign-off and acceptance testing.
  • [ ] Agree a cut-off date so records created mid-migration are not lost.
  • [ ] Plan a reconciliation pass comparing record counts and spot-checking samples after import.

Verifying the result: three different failures

> When something goes wrong after a migration, the cause is usually one of three distinct problems, and each needs a different fix.

Export failure. Records never left the old system. Symptom: the source file has fewer rows than your old ATS reports. Fix: re-run the export with the correct filters and date range.

Import failure. Records arrived in the file but did not load. Symptom: row counts in the new system fall short of the file. Fix: check the import log for rejected rows, usually caused by malformed dates, encoding issues or unmapped required fields.

Reconciliation failure. Records loaded but are wrong or duplicated. Symptom: counts match but spot checks show mismatched owners, lost notes or duplicate candidates. Fix: correct the mapping and re-import the affected subset.

A practical acceptance test: pick ten records you know well, including one with attachments and one with extensive notes, and confirm each looks right in the new system before you declare the migration complete.

Worked example: a mid-sized agency switching ATS

> Consider an agency with an export containing candidates, client companies, notes and resume files. Using the self-service route, the team imports candidates and companies through the Excel/CSV importer, attaches notes to the relevant records, then uploads resume files for parsing.

Suppose the source export has 12,000 candidate rows. After import, the new system shows 11,940. The gap of 60 is not automatically a failure: it may be duplicates merged during cleaning, or rows rejected for malformed data. The team checks the import log, confirms the 60 are accounted for, and only then signs off. Had they simply compared totals without reading the log, they would have had no way to tell a cleaning decision from a silent data loss.

This is why the 99% expectation is best treated as a planning figure rather than a guarantee: the useful question is not whether the number matches exactly, but whether every difference can be explained.

What you can do on this page

> The page offers several interactive elements. A video preview button opens a short walkthrough of the migration process. The FAQ accordion expands answers to common questions about security, timelines, data size, GDPR, ownership and cost. Testimonial controls let you move between customer quotes, and a demo request button connects you with a migration specialist.

If you are evaluating the service, the demo request is the most direct route to answers about your specific dataset, since pricing for managed migration depends on the volume of data involved.

Terms to confirm before committing

> Several commercial and compliance details depend on the provider you choose and should be confirmed in writing rather than assumed from a general description.

  • Managed migration pricing: {{OWNER_MIGRATION_PRICING_BASIS}}
  • Expected completion window: {{OWNER_MIGRATION_TIMELINE}}
  • Reported migration success rate and completed-migration count: {{OWNER_MIGRATION_SUCCESS_RATE}}
  • Data retention during and after migration: {{OWNER_DATA_RETENTION_POLICY}}
  • Security controls (hosting, encryption, backup cadence): {{OWNER_SECURITY_CONTROLS}}
  • GDPR compliance position and controller/processor responsibilities: {{OWNER_GDPR_COMPLIANCE}}
  • Data ownership and export rights: {{OWNER_DATA_OWNERSHIP_TERMS}}
  • Support response commitments: {{OWNER_SUPPORT_COMMITMENT}}
  • Contract duration and exit terms: {{OWNER_CONTRACT_TERMS}}

Treat any timeline or success figure you have seen elsewhere as a reference point to verify, not a commitment.

Self-service import vs managed migration
AspectSelf-service importManaged migration
CostFree, using the built-in importerPriced by volume of data to import
What it coversCandidates, companies/clients, notes; resumes via AI parserFull dataset handled by migration engineers
TimeDepends on your own effortTypically 10–12 business days
Best forTidy exports and teams with spare capacityLarge, mixed or messy exports
Who does the workYour teamProvider's data migration engineers

Route descriptions reflect the source platform's stated offering; confirm current pricing and timelines directly.

Frequently asked questions

Is my data secure while it is being migrated?

The source platform states that data it holds is stored in data centres managed by Amazon Web Services and encrypted with AES-256 bit encryption, which it describes as the global industry standard for internet data security, and that regular backups are taken. How the export file itself travels between systems is a separate question worth confirming with the migration team.

How long does a managed migration take?

The source platform states that most projects are completed within 10–12 business days, and reports a 100% success rate across 250+ migrations. Treat both as the source platform's historical claims to verify rather than a guarantee for your dataset.

How much data can be migrated, and in what formats?

The source platform states that any size of data is accepted, in CSV, XLS, XLSX, .BAK, .BACPAC, .SQL and database formats, and that 99% of your data will be cleaned and transferred. Plan a reconciliation pass for the remainder.

Is the platform GDPR compliant?

The source platform states that it is GDPR compliant and supplies tools intended to help your organisation meet its own GDPR obligations. Your own responsibilities as a data controller still apply.

Who owns the data once it is in the new system?

The source platform states that you own all the data you put in, that its role is to help you manage it, and that you can extract your data at any time. Confirm the equivalent ownership and export terms with whichever provider you choose.

What does migration cost?

There are two routes. The free option uses the built-in Excel/CSV importer for candidates, companies/clients and notes, plus an AI resume parser for CVs. The paid option uses the provider's migration engineers, priced according to the volume of data and typically taking 10–12 business days. Confirm current pricing directly.

If record counts don't match after import, is that always data loss?

No. A shortfall can come from duplicates merged during cleaning or rows rejected for malformed data. Check the import log and confirm each difference is explained before treating it as loss.