ATS Data Migration: Switch ATS Without Losing Candidate History

The thing that keeps most firms on software they have outgrown is not price. It is the database.

Years of records, notes and client history, plus the accumulated knowledge of people who have since left. Nobody wants to be the person who signed off on the migration that lost it.

That fear is reasonable, but it is aimed at the wrong half of the problem. Structured data moves fine. The context around it is what quietly does not.

TL;DR

• Structured data migrates reliably. Notes, call logs, emails and attachments are where loss happens.
• 71% of placements come from candidates already in the database, so context is the asset you are actually moving.
• Most quoted migration timelines cover structured data only. Ask explicitly.
• Clean before you move, test on a subset, and keep the old system readable for a defined period.

What actually gets migrated, and what gets lost

Migrations rarely fail loudly. The records arrive and the meaning does not. It helps to think about your data in three tiers.

Tier one: moves reliably

Candidate and contact records, company records, job and search records, pipeline stages, dates, owners and most custom fields. Any competent migration handles these. If a vendor cannot, walk away.

Tier two: moves only if you ask

Notes, activity history, email threads, call logs, attachments, tags and the audit trail. This is where vendors differ enormously, and where quoted timelines start to mean different things.

Tier three: usually rebuilt, not migrated

Automation and workflow rules, saved searches, report definitions, permissions and integrations. Rebuilding these is normal. Assuming they will transfer is not.

The expensive mistake is treating tier two like tier three: a nice-to-have you can recreate later. You cannot. Nobody is going to re-listen to four years of calls.

Why notes and call logs are the real risk

A record with a name, a title and a phone number is a lead. A record with three years of notes is an asset.

Across 2,100+ firms, our benchmark study found that 71% of placements come from candidates already in the CRM before the job opened (Source: The Economics of Recruiting).

That only holds if consultants can still tell why those people matter:

• The 2023 note saying they will not relocate but would revisit after their bonus vests
• The call log showing you have spoken four times, and who introduced you
• The reason they were rejected last time, which was about the client, not them
• The fact that they are now a hiring contact somewhere you want to sell

None of that sits in a structured field.

The risk is easy to miss because the loss is invisible at go-live. Record counts match. Reports run. It surfaces eleven months later, when a consultant finds a bare record and restarts a relationship from nothing.

The ATS data migration checklist

The sequence matters more than any single step, because the costly mistakes come from cleaning after the move instead of before.

1. Export a sample before you sign. Check whether notes and activity come out as usable data or an unreadable blob.
2. Decide what you are not migrating. Dead records, opt-outs, duplicate companies. Migration is the one moment when deleting is easy.
3. Deduplicate in the old system. Merging after the move means doing it twice.
4. Map every field, including custom ones. Decide for each whether it maps, merges into notes, or is dropped deliberately.
5. Agree in writing how unstructured data is handled. Attached to the right record, dumped into one field, or left behind. These are very different outcomes.
6. Test on a representative subset. Not your cleanest hundred records. Include the messy ones, your oldest, and a candidate who is also a contact.
7. Validate against checks defined in advance. Counts by type, twenty spot-checks you know by heart, notes on the right records, attachments intact.
8. Run parallel with an end date. Two to four weeks. Without a deadline it becomes permanent and your data diverges.
9. Freeze, cut over, keep the old system readable. Most firms need something from it in the first two months.
10. Retrain on search, not navigation. Consultants find the buttons themselves. They will not learn a new search behaviour unprompted.

Questions to ask any ATS vendor

Does migration include notes, call logs and attachments?

The single most important question. Quoted timelines frequently cover structured records only. Get the distinction in writing.

Can we see a migration that included unstructured data?

Not a reference call about how helpful the team was. A specific example of notes and activity landing correctly.

Who actually does the work?

Whether the vendor runs it or hands you a template changes your internal cost substantially.

What is the rollback plan?

Including what happens if the test migration fails validation, and how long you keep access to the old system.

If you are still shortlisting, our guides to the best applicant tracking software, ATS versus CRM and the recruitment tech stack cover the wider evaluation.

How Recruiterflow handles it

Recruiterflow migrates notes, call logs, activity history and attachments alongside structured records, with a dedicated team that runs the work rather than handing you a template. Process, supported platforms and timelines are in our guide to migrating from your current ATS.

The bigger point is what happens next. If notes are thin because recruiters have to type them after the call, a new system inherits the same habit. AIRA Notetaker joins the call, transcribes it and updates the record from what was said, so the database you migrate becomes the worst version you will ever have.

Frequently asked questions

How long does ATS data migration usually take?

Two to six weeks from kickoff to cutover for most firms, depending on database size, custom fields and how much cleaning is needed. Much shorter estimates are often quoting structured data only.

What data is typically lost when switching ATS?

Structured records almost always survive. The usual losses are notes, call logs, email threads, attachments, activity timelines, saved searches, report definitions and workflow configuration.

Can you migrate candidate notes and call logs to a new ATS?

Yes, depending on the vendor and what your current system exports. Ask whether notes arrive attached to the correct candidate and job, or as text in a single field, because that decides whether they stay searchable.

What should a data migration checklist include?

A sample export reviewed before signing, exclusion rules, deduplication at source, a full field map, a written agreement on unstructured data, a test migration, predefined validation checks, parallel running with an end date, and retraining on search.

How do you clean candidate data before migrating?

Work in the old system. Merge duplicates, standardise company names, fix malformed emails, archive records beyond your retention period and remove anyone who asked to be deleted. Cleaning first means doing it once.

What should you ask a vendor about their migration process?

Whether unstructured data is included, whether they can show a completed example, who performs the work, what happens if validation fails, and the rollback and retention terms.

The database is the business

Delaying a migration trades another year on a system that no longer fits against a few weeks of disruption. That is only rational if the risk is as large as it feels.

For structured data it is not. For your notes and call history it can be, and that risk is managed by asking one specific question early rather than waiting another year.

If you want to know exactly what would come across from your current system, book a demo and bring a sample export.