ATS Migration Checklist14 steps to switch without losing data, integrations, or candidate trust
Most ATS migrations get sold as a one-click import. In reality, the data part is easy. The hard part is everything around it: careers pages, integrations, scorecards, automation logic, signed offers, and the EEO records you legally have to keep. This is the checklist I wish someone handed me the first time I helped a team switch.
A realistic 8-week timeline for most teams under 200 employees
Discovery
Week 1
Data audit
Week 2
Configuration
Week 3-4
Parallel run
Week 5-6
Cutover
Week 7
Stabilize
Week 8+
Teams switch ATS for predictable reasons. The old one is too expensive, too slow to set up new workflows, weak on AI, or stuck on a price model that punishes growth. If you have read our breakdowns of Workable alternatives, Greenhouse alternatives, or the Ashby alternatives roundup, you already know the answer to "which ATS." This guide is about the migration itself.
Two things to set straight before we get into the steps. First, SHRM and most HR technology analysts treat ATS migration as a 6 to 12 week project for mid-market teams. If your vendor pitches "we will migrate you in a week," ask them which features they plan to skip. Second, the EEOC recordkeeping rules require employers to keep selection records for at least one year. That obligation does not pause because you are switching tools.
The plan below has 14 steps grouped into four phases. Pick a phase owner, put it on a shared timeline, and stop trying to do this on the side of someone's desk.
Phase 1: Discovery and audit (Week 1 to 2)
This phase costs nothing and saves everything. Most failed migrations skip it. The team buys a tool, kicks off implementation, and discovers in week six that the new ATS does not support a workflow the legacy system was holding together with duct tape and a custom field.
Step 1
Document what the current ATS actually does
Not what it was supposed to do when you bought it. What it is doing in production right now. List every active workflow, every integration that fires, every report someone opens weekly, every email template, every offer letter version. If your team uses Zapier or a homegrown script to patch around a missing feature, write that down too. Those are the workflows that disappear silently in a migration.
A simple way to do this: pull a list of every user in the legacy ATS and ask three questions. What do you log in to do? What is annoying? What would break if this disappeared tomorrow? You will get 80 percent of the truth in two days.
Step 2
Inventory every integration and data feed
Job boards. HRIS. Calendar. Email. Slack. Background check provider. Assessment tools. Video interview tools. Notetakers. SSO. Data warehouse exports. Each one of these will need a rebuild or a swap. Get the vendor name, the integration type (native, API, Zapier, manual), and the person who owns it.
A surprising number of teams find five or six tools in this list that they are still paying for and nobody uses. Migration is a good moment to cancel them.
Step 3
Identify your data must-haves before talking to vendors
This is where most teams under-prepare. They assume any modern ATS imports candidate data cleanly. Some do. Plenty do not. The list below is the floor. If a vendor cannot commit to all of it in writing, that is a real warning sign.
The non-negotiables. Get these in writing from your new vendor.
Candidate profiles
Resumes and attachments
Stage history with timestamps
Interview feedback and scorecards
Communication history (email, SMS)
Offer documents and signatures
Source attribution and UTM data
EEO and compliance records
Step 4
Confirm what data you can legally archive vs delete
The EEOC requires employers to keep selection records for at least one year after the action they relate to, with longer requirements for federal contractors. State laws often go further. GDPR adds candidate consent and right-to-be-forgotten obligations for anyone you have ever interviewed in the EU.
Talk to your legal or HR lead before you assume "we are deleting old applications" is a valid answer. The safest path is to archive everything subject to a retention rule and only migrate what you actually need to operate.
Phase 2: Vendor and configuration (Week 3 to 4)
By now you have picked your new ATS. The work shifts from inventory to setup. This phase decides whether you ship something usable on day one or something that needs three months of cleanup.
Step 5
Get the data migration plan in writing
Ask for a written migration scope document from the new vendor. It should specify which fields will be mapped, which will not, how custom fields are handled, what happens to attachments, and how stage history is preserved. If they will not give you that document, you have your answer about what kind of partner they will be in six months.
Then test it. Most vendors will run a sample export and import of 100 to 500 candidate records before contract signature. If they refuse, push back hard. A vendor that does not believe in their own importer is asking you to take on the risk.
Green flags in a vendor
- Vendor provides a written data migration plan
- Sample export and re-import test before signing
- Named migration owner with weekly check-ins
- Defined rollback path during parallel period
Red flags in a vendor
- Vague promise that the importer handles everything
- No support for stage history or interview notes
- Hard cutover with no parallel run window
- Integration rebuild left for the customer to do solo
Step 6
Rebuild stages and scorecards before importing candidates
If you import candidates into default stages and then rename them later, you will lose stage history. Build out your pipeline templates first. Then map every legacy stage to a new stage in a spreadsheet. Have the recruiter who actually uses the pipeline check it. They will spot the gaps a project manager will not.
This is also the moment to rebuild scorecards properly. If you are coming from a system where interviewers wrote feedback in a free-text box, use the switch as a forcing function to introduce structured evaluation. Our interview scorecard guide walks through what a defensible scorecard actually looks like.
Step 7
Set up SSO, permissions, and access groups
Most teams get this wrong by being too permissive. Default to least-privilege roles. Hiring managers see their own jobs. Coordinators get scheduling rights but not pipeline edits. Recruiters get full access on assigned reqs. Admins are a short list. Wire it to SSO from day one if your plan supports it. Cleaning up access after launch is twice the work.
Step 8
Connect HRIS, calendar, and email integrations
Get these working in the new system before you import any data. Two reasons. First, the test data you put in during sandbox lets you catch broken sync logic without polluting candidate records. Second, calendar and email auth tokens often need IT involvement, and IT calendars do not run on your timeline. Start early.
Phase 3: Data import and parallel run (Week 5 to 6)
The riskiest phase. Two systems are live. Recruiters are confused. Candidates are still applying. You need a clear rule for which system is the source of truth for which records.
Step 9
Run a pilot import with one team or one job family
Resist the urge to import everything at once. Pick one business unit, one job family, or even one open requisition. Import the candidates. Walk through every record with the recruiter who owns those candidates. Check the resume opens correctly. Check the stage history is intact. Check that interviewer notes are not flattened into one wall of text. Check that source attribution survived.
If anything breaks in the pilot, fix the mapping before the full import. Re-importing a few hundred records is a Friday afternoon. Re-importing fifty thousand is a project.
Step 10
Define the parallel-run rules clearly
During parallel run, draw a hard line. New requisitions open after the cutoff date go into the new ATS only. Active candidates stay in the system they started in until the role closes or they reach a defined transfer point. Decide which system runs offers and signatures. Decide which system owns the careers page during this period.
Write these rules down on one page and pin it in your recruiter Slack channel. Without that, half your team will be in the wrong system at any given moment.
Step 11
Train people on the workflows they actually do
Generic ATS training is a waste of everyone's time. Each role has its own three or four workflows. Recruiters do sourcing, screening, scheduling, and feedback collection. Hiring managers approve, review, and decide. Coordinators schedule and send offers. Build a 30-minute walkthrough for each of those persona-based workflows. Record it. Make people watch the relevant one before they touch the system.
Adoption problems on a new ATS are almost never about the tool. They are about people getting dumped into a generic training session and never being shown the path through their actual job.
Step 12
Validate a sample of migrated records side by side
Pick 50 candidates at random from the legacy system. Open each one in both systems. Compare every field you care about. Resume opens, stage history, feedback notes, source, EEO data, communication threads. This audit catches the silent losses that vendor reports will not. Plan for two or three rounds of corrections after the first one.
Switching to an AI-native ATS?
Prepzo is built for teams who want clean migrations, structured scorecards, and AI screening that actually moves the needle. Free trial, unlimited users, no per-seat pricing.
Try Prepzo freePhase 4: Cutover and stabilize (Week 7 to 8+)
Cutover is the day the legacy system goes read-only and the new one becomes the source of truth. Done well, it is a quiet Tuesday. Done poorly, it is a week of broken integrations and angry hiring managers.
Step 13
Run the cutover with a written runbook
Pick a low-traffic day. Friday afternoon is traditional. Walk through every cutover task on a checklist. Final export from the legacy ATS. DNS change for careers page. Re-publish jobs to job boards from the new system. Reconnect calendar and email per user. Send a one-paragraph note to all active candidates explaining what they will see change.
Set the legacy system to read-only access. Do not delete it for at least 90 days. You will reference it for one-off lookups for longer than you expect.
Cutover day, in order
Final data export from legacy ATS taken
New careers page domain swap verified
Job board integrations republished from new ATS
Email and calendar sync reconnected for each user
Active candidates notified of any system change
Legacy system set to read-only access
Step 14
Stabilize for 30 days before celebrating
The first month after cutover is where adoption is won or lost. Run a 15-minute weekly office hours session. Watch your pipeline metrics. If your time to hire or interview scheduling rate degrades, find out why fast. Often it is one missing automation or one permission gap, not the tool itself.
Hold a retro at day 30. What broke, what saved time, what people miss from the old system. The retro is also where you decide which "we will add that later" items from setup actually need to ship.
What about the careers page?
Easy to forget until launch day. The careers page is usually hosted by or wired to your ATS. Switching ATS means rebuilding it or repointing it. Two paths.
Path one: use the new ATS's hosted careers pages. Fastest. Limited brand control. Good for teams under 50 hires per year.
Path two: use the new ATS's embed widget or API on your existing marketing site. More control. Requires a developer hour or two. Better for brand-heavy teams. Whichever route you pick, validate that Google Jobs structured data, Indeed feeds, and LinkedIn job slot integrations all repoint cleanly. Capterra reviews of ATS migrations consistently call out this step as the most under-estimated.
The mistakes that cost teams the most
Assuming the importer handles everything
It usually handles 70 to 85 percent. The other 15 percent is custom fields, interview notes, and source attribution. That is where the silent losses live.
Migrating in the middle of a hiring crunch
If you are about to close ten roles this quarter, do not migrate now. The dip in productivity during weeks 5 to 8 is real. Pick a calmer window or accept the slowdown.
No named project owner
If everyone is in charge, nobody is. Pick one person on the people team and give them 30 to 40 percent of their week back for the duration. Without that, the project becomes side-of-desk and stretches to three months.
Underestimating training time for hiring managers
Recruiters live in the ATS. Hiring managers visit it twice a week. They will need a clear one-pager and a short Loom video, not a generic vendor webinar.
Forgetting about offer letter templates
Approval routes, signature workflows, country-specific clauses. Easy to assume they will move. They do not. Rebuild them in week 3, not week 7.
A note on AI-native ATS migrations
If you are moving from a legacy ATS to an AI-native one, the migration has one extra dimension: deciding which AI workflows you want on by default. Auto-screening, candidate matching, interview note capture, pipeline summaries. Each one is a meaningful change to how recruiters spend their time.
My recommendation: turn on AI screening from day one for high-volume roles, where the time savings are obvious. Keep AI-driven candidate matching off until you have rebuilt your scorecards, otherwise the model will optimize against weak signals. Turn on interview notetaking after the first two weeks, once interviewers have a chance to set expectations with candidates about recording consent.
For a deeper view on how AI changes the workflow, our AI hiring playbook and the breakdown of why traditional ATS analytics are broken cover the reasoning.
Plan your ATS migration with a partner who has done it
Prepzo's onboarding team has helped teams migrate from Greenhouse, Workable, Ashby, JazzHR, and BambooHR. Talk to us before you sign anywhere.
Try Prepzo freeFrequently Asked Questions
How long does an ATS migration usually take?
For a small or mid-sized team with under 10,000 candidate records, 4 to 8 weeks is realistic from kickoff to full cutover. Larger teams with complex integrations, multiple business units, or heavy custom workflows often need 3 to 6 months. The variable is rarely the import. It is the rebuild of careers pages, scorecards, automations, and offer templates.
What candidate data should we export before switching ATS?
At minimum: candidate profiles, resumes, application history, stage history with timestamps, interview feedback, scorecards, communication history, offer documents, source attribution, and EEO data. Anything tied to a hire decision should be exported and retained in line with EEOC record-keeping rules, which generally require keeping selection records for at least one year.
Can we run two ATS systems at the same time during migration?
Yes, and most teams should. A 2 to 4 week parallel period lets you keep active pipelines moving in the old system while training the team and validating data in the new one. Freeze new requisitions in the legacy system after a clear cutoff date so you do not end up reconciling two sources of truth forever.
What is the biggest risk in an ATS migration?
Silent data loss. Vendors love to say their importer handles everything. In practice, custom fields, stage history, and interviewer notes often drop or get flattened into a single text blob. The fix is to audit a sample of 50 migrated candidate records side by side with the source before you sign off on the cutover.
Should we migrate historical data or start fresh?
Migrate active candidates, recent hires, and at least 12 months of historical records for compliance. Beyond that, weigh the cost. For most teams, a clean archive of older data plus a focused import of the last 18 months works better than dragging years of stale records into the new system.
How do we keep candidates from falling through the cracks?
Map every active stage in the legacy system to a corresponding stage in the new one before import. Tell candidates in active pipelines what is happening, even briefly. Most ghost cases during migrations come from rejection emails that never sent or interview invites that bounced because the integration changed mid-flight.
Resources & Further Reading
Related Guides
- What Is an Applicant Tracking System?
The fundamentals before you switch tools
- AI Hiring Playbook for an AI-Native ATS
How the workflow changes when AI is built in
- Best Greenhouse Alternatives in 2026
If you are switching off Greenhouse specifically
- 15 Recruitment Metrics & KPIs to Track
How to know if the new ATS is actually working
External Sources
- EEOC Recordkeeping Requirements
Federal rules on retaining selection records
- SHRM Technology Section
HR technology selection benchmarks
- Capterra ATS Software Reviews
Real customer migration experiences
- Google re:Work Structured Interviewing
Build better scorecards before you import
