Duplicate contacts in a CRM: why they appear and how to merge them

Why duplicates appear

• Several channels: the same person fills in a website form, answers an ad lead form and later calls. Each channel creates its own record.
• Different formats: +380 67 123 45 67, 0671234567 and 067-123-45-67 are the same number but not the same text. The same happens with names written in Latin and Cyrillic, or with and without a middle name.
• Imports: a spreadsheet from an old system or a trade fair is imported on top of records that already exist.
• Typing instead of searching: a manager creates a new card because finding the existing one seems slower.
• People change: a new email, a new job, a second phone number.

What duplicates cost

The history of a client is split between records, so nobody sees the whole picture: one card has the orders, another the last conversation. The client gets contacted twice, sometimes by two different managers. Reports count one customer as two and a repeat buyer as two new ones. Consent becomes unreliable: if a person unsubscribed in one record, the other record may still be on a mailing list.

How to find duplicates

Look for matches in order of reliability:

• Phone number, compared in one format, for example the international form with the country code. For private clients this is usually the most reliable key.
• Email address, compared in lower case and without spaces.
• Name plus company or city, for records without phone or email. These are only candidates and need to be checked by hand.
• Company name for organizations, including variants with and without the legal form, such as LLC, GmbH or ТОВ.

Shared numbers are a trap. An office switchboard or a family phone can belong to several real people. A match is a signal to check, not an automatic verdict.

How to merge safely, step by step

1. Before a large clean-up, export the records so that you can check later what was there.
2. Review each candidate pair and confirm that it really is the same person.
3. Choose the main record: usually the one with the most history, linked orders or the most complete data.
4. Compare the fields and keep the correct values: the current phone, the confirmed email, the right spelling of the name.
5. Move notes, tags, addresses, orders and other linked records into the main record.
6. Check consent and communication preferences: if any of the records says the person unsubscribed, the merged record must respect that.
7. Remove the other records, and tell the team which card is now the main one if they worked with the others.

Worked example (illustrative numbers)

A company imports 1,200 contacts from an old spreadsheet into a CRM that already holds 3,000. After the import, a check by phone number finds 140 pairs with the same number. On review, 118 pairs turn out to be the same people entered twice, while 22 are colleagues sharing an office phone and stay separate.

Of the 118 real duplicates, 30 have orders on both records. After merging, each of these clients has one card with the full order history, and 30 customers who looked like first-time buyers in reports turn out to be repeat customers. During the review the team also finds 9 people who had unsubscribed in one record but were still on a mailing list through the other. Without the merge, these people would have kept receiving messages they asked not to get.

How to prevent new duplicates

• Search before creating a card: by phone first, then by name.
• Store phone numbers in one format with the country code, so that matches can be found.
• Bring every channel into one place instead of letting each one create records on its own.
• Check every import for duplicates, both in the file and against existing records.
• Review likely duplicates regularly, for example once a month, instead of once a year.

How this is done in OrgLines

OrgLines flags likely duplicates: contacts that share a phone number and organizations with the same name. In the merge dialog you choose the main record. The tags, notes, addresses, orders and customer or supplier status of the other contacts are merged into it, and the remaining records are deleted. When you merge organizations, their contacts, deals, projects and orders move to the main organization in the same way.

Some duplicates are prevented before they reach the contact list: a quiz on your website merges answers with the same phone number into one lead instead of creating several. Existing contacts come in through the import wizard, where you upload a file, map its columns, validate and import; duplicates found after the import can then be merged. Because the other records are deleted after a merge, review each pair before you confirm.

Common mistakes

• Merging automatically on the name alone, which joins different people with common names.
• Treating a shared office or family phone as proof of a duplicate.
• Choosing the newest record as the main one instead of the one with the history.
• Losing an unsubscribe or consent flag during the merge.
• Deleting duplicates instead of merging them, which throws away notes and orders.
• Cleaning up once and never again, so that duplicates are back within months.

Checklist

• Phone numbers stored in one format with the country code.
• Candidates found by phone, email and name.
• Each pair reviewed by a person before merging.
• Main record chosen by history, not by date.
• Notes, tags, addresses and orders moved to the main record.
• Consent and unsubscribe flags checked.
• Regular review of duplicates scheduled.

Try it free in OrgLines