An Applicant Tracking System that creates a second candidate record whenever a job-board apply, a referral or a re-application arrives — even when email, phone or LinkedIn already matches — splits history, notes, consent and stage state across two profiles. Require merge (or an explicit keep-separate with a recorded reason) so every match lands on one honest record, without vanity metrics or promising that artificial intelligence replaces recruiters.
This matters for recruitment agencies, staffing, secondment, executive search and in-house recruitment: duplicates are separate from next-action dates, from standardized rejection reasons and from first-review deadlines. If two records share one person, desks chase the wrong history, consent is incomplete and stage state lies.
Why duplicate candidate records kill pipeline truth
Allowing a second profile “just to be safe” looks cautious and still loses control. Typical failure modes:
- Job-board, referral and re-application intakes create new candidates despite matching email, phone or LinkedIn
- Notes, interview outcomes and client feedback sit on the old record while the new one owns the live stage
- Consent and communication preferences diverge across two profiles for the same person
- Desks waste time reconciling history by hand or treat one profile as “dead”
A candidate and client CRM beside the pipeline only helps when the candidate side is one person, one record. Merge is how that stays true.
What merge (and keep-separate) means
Define it clearly in the Applicant Tracking System:
- Match signals — email, phone and LinkedIn URL as primary keys; secondary signals (name + company) may suggest review but must not auto-merge
- Merge — combine history, notes, applications, consent and stage-relevant data into one surviving candidate record; retire or redirect the duplicate
- Keep-separate — an explicit human decision with a short recorded reason (for example two people sharing a family email, or a known false match), never silent creation of a second profile
- What it is not — a vanity “dedupe score” dashboard, automatic discard of history, or artificial intelligence deciding identity without a human confirm
If duplicates stay optional, desks invent workarounds. Merge only works when every high-confidence match forces a decision: merge or keep-separate with reason.
How to require merge in the Applicant Tracking System
Process beats goodwill. In the Applicant Tracking System:
- On intake, surface possible duplicates when email, phone or LinkedIn match an existing candidate
- Block creating a second record until the desk merges or records keep-separate with reason
- Preserve notes, applications, consent and stage history on the surviving record after merge
- Show a queue of unresolved duplicate suggestions so matches do not die in personal inboxes
Pair this with a single vacancy owner so someone owns clearing merge decisions on open roles — while merge itself protects one identity across the whole desk.
Alerts and desk review without vanity metrics
Alert on unresolved high-confidence duplicate matches and on second records created without a keep-separate reason. Do not invent “dedupe rates” or cleanliness scorecards nobody trusts. In the next desk meeting, review how many merges were completed, how many keep-separate reasons were used and whether false matches need a tighter rule. That review is quieter than a data-cleanup project and more honest than vanity charts.
Where artificial intelligence may help — and where it must not
Artificial intelligence may suggest possible duplicates for a human to confirm, or rank match confidence so the queue is easier to clear. It must stay secondary. Artificial intelligence does not replace recruiters, does not auto-merge without human confirmation, and does not invent identity. Humans confirm the person. The Applicant Tracking System stores the match signals, the merge or keep-separate decision, the reason and the surviving record.
How merge sits next to next-action dates, rejection reasons, CRM and first review
Duplicate merge complements other Applicant Tracking System habits:
- Next-action dates belong on one open candidate–vacancy; two records mean two clocks and neither is trusted
- Standardized rejection reasons only show patterns when past outcomes sit on the same person
- Candidate and client CRM beside the pipeline needs one candidate identity so relationship history is complete
- A first-review deadline still applies on intake — merge first so the review happens on the right record
- A single vacancy owner makes clearing merge queues someone’s job on open roles
Without merge discipline, the pipeline looks busy while the same person lives twice. With it, every match has a recorded decision until one record owns the truth.
A practical week to introduce merge discipline
One week in the Applicant Tracking System:
- List open candidates that share email, phone or LinkedIn with another record
- Merge each confirmed duplicate, preserving notes, consent and applications on the survivor
- Record keep-separate with reason wherever the match is false
- Turn on blocking of second-record creation on high-confidence matches
- In the next desk meeting, review unresolved suggestions and keep-separate reasons
That habit is quieter than a master-data project. It is also how history, consent and stage state stop splitting across two profiles for one person.
Merge every duplicate candidate — then move with intent
If your Applicant Tracking System allows a second record whenever intake looks “new,” desks will chase split history while candidates and clients see inconsistent contact. Require merge or keep-separate with reason on email, phone and LinkedIn matches. Keep artificial intelligence as a suggester only. Then explore the product or book a demo — we walk through BoldATS duplicate-merge discipline in your segment.