Run a duplicate detection sweep | Nynch Wiki Skip to content
Use cases Product Pricing Cohort Strategy Day Resources Sign in Talk to the founder
Home / Wiki / Contacts & Data Quality

Run a duplicate detection sweep

You'll learn how to: Trigger a fresh duplicate detection scan across people in your workspace.

Time: A few seconds to start. Results usually return within a minute for typical workspaces.

Prereqs: You can open Record checks or the Possible duplicates page.

When to use

Run a manual sweep when:

  • You just finished a big import (CSV, LinkedIn export, Gmail history).
  • You merged a teammate's people into your workspace.
  • You suspect duplicates that are not on the Possible Duplicates page yet.
  • You changed email or LinkedIn values in bulk and want to re-check.

Steps

  1. Open Possible duplicates:
    • From Record checks, click Possible duplicates, or
    • Go to /possible-duplicates while signed in.
  2. Click Run scan.
  3. Wait for the toast confirming how many pairs were written.
  4. Review pairs sorted by match confidence.
  5. For each pair: Review merge (three-column field review) or Ignore.

What the sweep checks

  • Email address (exact match), certain identity.
  • LinkedIn URL (exact match after normalisation), certain identity.
  • Strong possibles, similar name plus company and/or title signals.

Pure name-only weak matches stay out of the default review queue so the inbox does not flood with noise.

Each candidate pair gets a confidence score and plain-English match reasons (Same LinkedIn, Same email, Similar name, and so on).

After the sweep

  • Review merge opens a three-column review: Record A, Record B, and Final. You choose the survivor and conflicting field values, then confirm.
  • Ignore dismisses that pair so it does not reappear until you intentionally re-scan after identity fields change (dismissed pairs are not reopened by the scanner).

See Merge duplicate contacts for bulk merge from the People table.

If something goes wrong

  • Symptom: "Scan never run" empty state. → Fix: Click Run scan once. The page will show last-checked time after a successful run.
  • Symptom: "Sweep finished but no new pairs appeared." → Fix: No strong duplicates were detected, or pairs were already dismissed/merged. This can be healthy.
  • Symptom: "Same pair keeps reappearing after I dismiss it." → Fix: Dismissed pairs stay dismissed. If a new pair appears, the underlying people or identity fields changed.
  • Symptom: "Sweep takes forever." → Fix: Very large workspaces scan a recent slice of people first. Re-run after large imports.

Step 1: Open the relevant screen

Open Possible duplicates:.

Step 2: Complete the action

Click Run scan.

FAQ

Q: How do I run a duplicate detection sweep? Trigger a fresh duplicate detection scan across people in your workspace.

Q: What do I need before I start? You can open Record checks or the Possible duplicates page.

Q: How long does this take? About A few seconds to start. Results usually return within a minute for typical workspaces.

Q: Scan never run" empty state? Click Run scan once. The page will show last-checked time after a successful run.

Q: Sweep finished but no new pairs appeared? No strong duplicates were detected, or pairs were already dismissed/merged. This can be healthy.

Related: Merge duplicate contacts | CSV import step by step.