FAQ — Duplicate contacts
Duplicate contacts are inevitable. A customer emails from a personal address, then calls from a number not on file, then messages on WhatsApp. Each new identifier creates a new contact. Over time, your CRM accumulates duplicates. Here’s how to deal with them.
Why do duplicates exist at all?
Atender creates a contact when a new identifier (email, phone, channel ID) shows up. If a customer emails from john@example.com and you don’t yet have a contact with that email, a new contact is created.
When the same person later contacts you from a different identifier — john.smith@personal.com, or a phone Atender has never seen — Atender doesn’t know it’s the same person. It creates a new contact.
This is the right default. The alternative — auto-merging contacts based on name similarity — would lead to merges between people who share a name but aren’t actually the same.
Why doesn’t Atender warn me about duplicates automatically?
Atender doesn’t show “possible duplicate” prompts based on name or email similarity. The product doesn’t try to guess — duplicate detection is your call.
The reason: false positives in auto-merge are very expensive. Merging two contacts who share a name but aren’t actually the same person mixes their conversation histories, addresses, and custom field values into a single record. Recovery means recreating both records by hand. Better to let humans confirm.
How do I find duplicates?
A few patterns:
- Search by email — if two contacts come up for the same email, those are duplicates
- Search by phone — same logic
- Filter by linked organization — within one company, multiple records with similar names are usually the same person
- Sort by creation date and scan recent imports — bulk imports often produce duplicate clusters
- Open a contact and review its Conversations tab — if conversations clearly belong to a different person but are linked to this contact, you have a confused record
What happens when I merge duplicates?
Everything moves to the survivor:
- Conversations are reassigned to the surviving contact
- Email addresses, phones, and channel addresses union onto the survivor (deduplicated within type)
- Custom field values from the duplicates fill any empty fields on the survivor
- Attachments move over
- The non-survivor contacts are deleted
The survivor’s name, type, and customer tier are preserved. Pick the right survivor before merging — you can edit details after, but you can’t undo the merge.
See Merge contacts for the full procedure.
When should I NOT merge?
- Different people who share a name — even if their conversation history looks similar, two different humans are two different contacts
- A spouse and their partner using a shared family email — they’re separate people. Use a different identifier as the discriminator
- A personal account and a business account from the same person — usually you DO want to merge, but if the business and personal interactions need to stay distinct (different SLAs, different tiers), keep them separate
- Anything you’re not sure about — when in doubt, don’t merge. Two contacts can be re-merged later. A bad merge can’t be undone
What if I accidentally merged the wrong contacts?
Merging is one-way and there’s no built-in undo. To recover:
- Read through the survivor’s Conversations tab to identify which conversations belong to which “real” person
- Recreate the contact you accidentally merged away (manually, with the original identifiers)
- For each conversation that belongs to that person, re-attach it via the conversation’s contact field
This is slow and prone to error. Avoid it by being careful at merge time.
Why do I keep getting new duplicates from the same customer?
If a customer keeps showing up as a new contact, the customer is using identifiers Atender doesn’t recognize:
- A new email address
- A new phone number
- A different web-chat session (web-chat visitor IDs are session-based)
Add the new identifier to the existing contact rather than letting a new contact be created next time. Or, after the duplicate is created, merge into the original.
Do duplicates hurt anything?
Yes — in three ways:
- Agents see partial history — opening a duplicate shows only the conversations linked to that record, not all the customer’s interactions
- Reporting is fragmented — reports group by contact, so duplicate-customer activity is split across records
- Customer-tier and custom-field values can drift — VIP marker on one record, no marker on the other; agents see whichever they happened to open
Worth a quarterly cleanup pass for any tenant with significant CRM volume.