Back to all articlesContact identity

Check That a Contact Merge Preserved the Right History

Alactic Editorial10 min read

Prepared with AI assistance by Alactic Editorial. Examples identified as hypothetical are illustrative, not customer results.

A contact merge can leave one tidy record while breaking the history that makes it useful. An enquiry may disappear from the contact view, an old conversation may be attributed to the wrong person, or the same note may appear twice under different dates.

To verify an authorised merge, compare an evidence-backed list of expected history with what the surviving contact actually shows and links to. Check identity, conversation content, attribution, affiliations and linked enquiries separately. Record each difference as confirmed, explained or unresolved. A reduced contact count is only evidence that consolidation occurred—not that the right history survived.

The checklist below is designed for post-merge review. It does not cover choosing which field values should survive, reconciling a migration, or operating a particular merge tool. Use the records, views and approved evidence available in your workspace; do not assume that exports, audit trails, reversal or relationship repair are supported.

1. Establish what should have survived

Start with the authorised merge scope: which original contacts were included, which contact remains, who authorised the action and when it occurred. Preserve original record references where available. Names alone are weak anchors because names can be shared or changed.

Next, assemble the expected history from evidence that exists independently of the surviving contact view. Useful inputs may include a pre-merge review, an approved snapshot, original enquiry records or an existing merge log. Do not create additional copies of sensitive conversations merely to make the checklist convenient. A record reference and a short description may be sufficient.

If the merge has already happened and no baseline exists, reconstruct expectations from accessible underlying records and colleagues responsible for the work. Mark those expectations as reconstructed, and distinguish documented evidence from recollection. A colleague remembering an enquiry is a reason to investigate, not proof that the enquiry was attached to a particular source contact.

  • Identity anchors: original contact references and evidence that the records represented the same person.
  • History anchors: conversation, note, enquiry and relationship references, where available.
  • Meaningful context: dates, authors, participants, affiliation periods and the role of the contact in each linked item.
  • Known exceptions: items already missing, duplicated or uncertain before the merge.

This baseline prevents a common mistake: treating every post-merge irregularity as merge damage, including problems that already existed.

2. Verify identity before interpreting the timeline

First confirm that the surviving record represents the intended person. Matching names or a shared business phone number are not enough on their own. Compare the available identity evidence with the authorised scope, including whether an address or number belongs to an individual, a shared inbox or a team.

Look for signs that another person's history has entered the record: overlapping conversations with different participants, incompatible affiliations, or notes that explicitly name someone else. These are investigation signals, not automatic proof of an incorrect merge. An assistant, colleague or shared communications channel can explain some apparent contradictions.

Check the participant inside the interaction, not only the contact name displayed above it. A conversation containing several people should not become evidence that the surviving contact personally made every statement.

If identity remains doubtful, stop treating the combined history as a reliable basis for follow-up. Flag it for the authorised owner and avoid further consolidation while the issue is investigated. A complete-looking timeline is not a reason to overlook an identity conflict.

3. Reconcile conversations by content and attribution

A timeline count is a useful first signal, but it is not a reconciliation. A conversation can appear as one summary in one view and several events in another. Conversely, equal totals can hide one missing interaction and one duplicate.

For each expected conversation, inspect the underlying item where permitted and compare:

  • Participants: who actually sent and received the messages.
  • Timing: the original interaction date, distinguished from an import, edit or merge date.
  • Continuity: whether the visible exchange includes the expected beginning, response and outcome.
  • Context: whether attachments, related notes or referenced enquiries remain accessible where expected.
  • Attribution: whether the original author and channel are distinguishable from the person who performed the merge.

Do not interpret a date discrepancy until you know what the displayed date means. A list ordered by last activity can place an old conversation near the top without changing its original message dates.

Notes need a separate duplicate check. Compare content, author, original date and source reference if available. Identical wording may indicate duplicate copies, but it can also be a deliberately reused instruction. Avoid deleting a suspected duplicate until its provenance and links are understood. Deletion can remove the only copy that still carries useful context.

4. Check affiliations and enquiries as relationships

A contact's history is not limited to the timeline. The person may have worked for different organisations, represented different branches or participated in enquiries in different roles. A current organisation label does not establish the context of an older interaction.

For affiliations, compare the expected organisation, role and relevant period with the observed relationship. Where your workspace cannot represent dated affiliations, record that limitation in an approved place rather than silently treating the current employer as the employer for all past events.

For enquiries, follow the link and inspect the destination. Confirm that it is the intended enquiry and that the contact's role is still correct. A requester, billing contact and technical participant are not interchangeable. Also check whether the enquiry still points to an obsolete contact reference, if that relationship is visible to you.

A missing link does not necessarily mean the enquiry was deleted. It may still exist but no longer appear in the contact's related-items view. Equally, a visible enquiry title does not prove the relationship is correct. Separate the questions: does the item exist, can an authorised reviewer access it, and is it linked to the right person in the right role?

5. Use a filled post-merge reconciliation checklist

Fictional worked example: All names, references, dates, counts and findings in this checklist are hypothetical. They illustrate a recommended review design, not product functionality or customer results.

In this example, two records for Maya Chen were merged after an authorised identity review. Maya previously worked at North Quay Studio and now works at Cedar Works. The reviewer has an approved pre-merge inventory and can inspect the referenced conversations and enquiries.

Review header

  • Merge scope: hypothetical contacts C-184 and C-527 consolidated into C-184.
  • Authorisation: hypothetical operations lead approved the merge on 8 April; scope covers Maya Chen only.
  • Baseline: approved pre-merge inventory dated 8 April, plus original conversation and enquiry references.
  • Review scope: all inventoried conversations, notes, affiliations and linked enquiries; no sampling.
  • Known pre-existing issue: one old conversation already lacked its referenced attachment. This is recorded separately, not attributed to the merge.

Filled expected-versus-observed checks

  1. Identity — confirmed. Expected: both source contacts represent Maya Chen, with documented context for the former and current business addresses. Observed: the surviving record and reviewed interactions match that person; no unexplained third-party history was found in the reviewed scope. Evidence: approved identity review and participant details in the inventoried conversations.
  2. Conversation CV-021 — confirmed. Expected: a 12 February exchange about North Quay Studio's initial enquiry, including Maya's question and the team's response. Observed: both parts are accessible, with original participants and dates. Finding: no discrepancy.
  3. Conversation CV-063 — unresolved visibility discrepancy. Expected: a 19 March follow-up concerning the same enquiry. Observed: absent from the surviving contact timeline, but accessible through enquiry EN-204. Finding: the conversation exists; its contact association or display needs investigation. Owner: CRM administrator.
  4. Conversation CV-088 — confirmed. Expected: a 4 April Cedar Works discussion, separate from the North Quay work. Observed: accessible with the correct participants and linked to EN-318. Finding: no discrepancy.
  5. Note NT-014 — duplicate suspected. Expected: one note authored by the account coordinator on 13 February. Observed: two entries with matching text and author, one displaying the merge date. Finding: compare underlying references and date meanings before removing either entry. Owner: CRM administrator.
  6. Affiliation history — incomplete. Expected: North Quay Studio through February; Cedar Works from March. Observed: only Cedar Works is visible, while older enquiries still concern North Quay. Finding: historical affiliation context is not adequately represented. Owner: contact data steward, to determine a supported correction.
  7. Enquiry EN-204 — relationship not reconciled. Expected: Maya remains the requester for the North Quay enquiry. Observed: the enquiry exists, but its visible contact reference is C-527 rather than C-184. Finding: investigate whether this is a retained alias or a broken relationship; do not assume either. Owner: CRM administrator.
  8. Enquiry EN-318 — confirmed. Expected: Maya is the Cedar Works requester. Observed: enquiry links to C-184 with the expected role and context. Finding: no discrepancy.
  9. Known attachment gap — pre-existing exception. Expected: attachment unavailable before the merge. Observed: still unavailable. Finding: not evidence of new merge damage; retain the existing issue reference.

Recorded decision

Outcome: not yet verified. Identity checks passed, but conversation visibility, note duplication, affiliation context and an enquiry relationship remain unresolved. One surviving contact is therefore not sufficient for sign-off. The CRM administrator owns the technical investigation; the data steward owns the historical-context decision. Each finding needs an evidence reference, next action and agreed review date.

6. Investigate discrepancies without creating more damage

Classify each finding before attempting repair. Useful categories are missing item, missing relationship, duplicate representation, incorrect attribution, unexplained date difference and insufficient evidence. These categories lead to different actions.

For a missing conversation, first check whether it exists through another authorised route. For an apparent duplicate, establish whether there are two underlying records or two views of one record. For an incorrect affiliation, distinguish absent historical context from an actually incorrect relationship.

Ask an authorised administrator or support contact to investigate inaccessible evidence. Do not bypass permissions or broaden access merely to make reconciliation easier. If another reviewer can see an item, that may explain a visibility difference, but it does not prove the intended user's working view is correct.

Repair only through supported, authorised methods available in the workspace and current plan. Do not assume a merge can be reversed. An improvised second merge, manual recreation or deletion can obscure the original evidence and introduce new errors.

After a correction, repeat the affected checks and inspect nearby relationships for unintended changes. Record what changed and what evidence now supports closure. A temporary note explaining uncertainty can help colleagues avoid misinterpretation, but it is not a substitute for restoring a broken relationship.

7. Set a review depth you can defend

A full item-by-item review offers stronger coverage but takes more time. Sampling can reduce effort on a large history, yet it cannot establish that every interaction survived. State that limitation explicitly.

If sampling is necessary, select deliberately: include history from each source contact, the oldest and newest interactions, each represented affiliation, and conversations tied to active enquiries. Review identity conflicts and suspected misattribution directly rather than leaving them to chance.

Expand the review when a sample reveals a discrepancy. One broken enquiry relationship is a reason to inspect other enquiry relationships, not just repair the one that happened to be noticed. Similarly, duplicate notes associated with one source record justify checking other notes from that source.

Record both coverage and unresolved findings. Reporting that a review found no discrepancies is misleading if the reviewer examined only a small, unrepresentative part of the history.

8. Close on evidence, with a measurable next action

Use clear closure states: verified within the stated scope, verified with documented exceptions, or not yet verified. Documented exceptions should have an owner and an explicit decision about acceptable use; they should not become a convenient label for unexplained identity or attribution problems.

For the fictional example, closure requires resolving the four open findings and repeating the affected checks. The pre-existing attachment gap remains a separate documented issue. The final review record should make that distinction visible.

For your next authorised merge review, create one expected-versus-observed line per inventoried item. Record the number checked, confirmed, explained and unresolved, along with the scope covered. Assign every unresolved discrepancy an owner and a review date. Sign off only when the evidence supports the stated outcome—not when the contact list looks cleaner.

Contact identityCRM operationsData qualityMerge verification

Put the next conversation to work.

Explore Alactic Communications for WhatsApp conversations, campaigns and AI-assisted workflows. Choose the product and plan that fit your team.

Explore Communications