Skip to content

Record merge

Part of: ClientVerse Core (every client).

What it is

Merges duplicate records into one surviving record, reparenting everything attached to the losers rather than discarding it. Duplicates are unavoidable in a CRM fed by several systems and many people; merge is how the firm converges on one record per real-world person or organization without losing the history spread across the copies.

Who it's for

Admins, gated by CV_Merge_Records. This is deliberately narrow: merge is destructive and effectively irreversible.

How it works

  1. Identify the duplicates and choose which record survives.
  2. Choose the surviving field values where the records disagree.
  3. Merge.

Related records from every merged record — activities, list memberships, classification assignments, connections, experience holders, wall memberships — are reparented onto the survivor. The losing records are removed.

Reparenting is the whole point. A merge that keeps the right record but drops its related data has destroyed exactly what the merge was meant to preserve. Any related record left orphaned or deleted is a data-loss defect of the highest severity, not a cosmetic gap.

Before you merge

Merge is not reversible in any practical sense — there is no unmerge.

  • Confirm they are actually duplicates. Two people with the same name at the same company are not necessarily one person.
  • Look at what is attached to each side before choosing the survivor, so you know what should exist afterwards.
  • Note the counts. Activities, list memberships, connections — check them after the merge; the totals should be the sum of both sides, less genuine overlap.
  • Prefer the richer record as the survivor where the choice is otherwise arbitrary.

After you merge

Verify rather than assume:

  • Related-record counts match what you noted.
  • Rollups that included either record have recalculated correctly (see Rollups).
  • List and campaign memberships survived — a contact silently dropped from an active campaign audience is easy to miss and externally visible.

Preventing duplicates upstream

Merge is remediation. The upstream controls matter more:

  • Data quality monitoring surfaces records likely to be duplicates.
  • Data validation and enrichment reduce the malformed entries that become duplicates.
  • Integration configuration determines whether inbound syncs create duplicates in the first place — a feed matching on the wrong key will regenerate duplicates faster than anyone can merge them.

If merge volume is high and persistent, the fix is upstream, not more merging.

  • Admin Guide — data stewardship.
  • Rollups — verify recalculation after merging.
  • QA coverage: DAT-06, DAT-07 in core-data.md.