How to remove duplicate contacts in GoHighLevel

Step by step

1. Export the sub-account

   Contacts → select all → Export. Take a full CSV including custom fields, tags, and the date-created column. Keep this file untouched as your rollback.

2. Standardise before you match

   This is the step people skip and it wastes the whole exercise. Normalise casing, strip company suffixes, and convert every phone number to E.164 format. ‘07700 900123’ and ‘+447700900123’ are the same number, but no matching engine knows that until you tell it.

3. Extract the email domain

   Split every address into local part and domain. Matching on domain plus surname catches duplicates that exact-email matching never will — dave@acme.co and d.mills@acme.co are plausibly the same person at the same company.

4. Run fuzzy matching, not exact

   Match on a combination: surname similarity, email domain, and normalised phone. Any two of the three agreeing is usually a genuine duplicate. Flag rather than auto-merge where only one matches.

5. Merge, never delete

   For each cluster, build one surviving record that keeps the most complete value from every field. If one duplicate has the phone number and another has the job title, the survivor gets both. Log every merge.

6. Consolidate the tags

   Duplicates almost always carry tag variants — lead, Lead, new-lead, Lead-2024. Decide the canonical tag, map the variants to it, and write the taxonomy down before re-import.

7. Re-import and verify

   Import as an update against a unique identifier, not as new contacts, or you will create the exact problem you just fixed. Spot-check 50 records against the original export.

What GoHighLevel's Manage Duplicates tool does — and where it stops

GoHighLevel ships a built-in Manage Duplicates tool. It is genuinely useful and you should try it before paying anyone, including us. Open Contacts, click the three-dot menu, choose Manage Duplicates, then scan by email, phone or name.

Four documented constraints decide whether it is enough for your account. All four come from HighLevel's own support documentation:

Use the native tool when you have a few dozen obvious duplicates and an admin with a free afternoon. Work outside it when you have thousands of pairs, need an audit trail, or cannot risk an unreversible merge on live client data. Below is the method we use for the second case.

Why duplicates keep appearing in the first place

GHL's native deduplication compares email and phone as exact strings at the point of entry. It is fast and it is safe — it will never merge two people who happen to share a surname. But it means every one of these creates a new record:

• A contact fills a form with dave@acme.co having previously been imported as d.mills@acme.co
• A phone number is entered as (555) 014-2 where the existing record holds +15550142
• A CSV import runs without a unique identifier mapped
• Two sub-accounts hold the same lead and get consolidated later

None of these are bugs. They are the predictable result of exact matching applied to human data entry.

How to stop it recurring

Cleaning is the easy half. Preventing recurrence takes three habits:

1. Map a unique identifier on every import. Email is usually the practical choice. An import without one is a duplicate generator.
2. Lock down tag creation. If everyone can invent tags, everyone will. Publish the taxonomy and review it monthly.
3. Normalise phone numbers at entry. A required-format field on your forms costs ten minutes and prevents a category of duplicate permanently.

Contact data still decays at roughly 22–25% a year regardless, so plan a verification pass each quarter.