How to Stop Duplicates From Getting Into Your CRM
Find where your duplicates come from, close the gaps you can, and merge the rest automatically.
Duplicates keep arriving after a cleanup, because forms, imports, integrations and reps keep creating records. Stopping them takes three layers: fix the sources you can, keep your CRM's own duplicate checks on, and use Dedupely to merge the duplicates that get past both, soon after they reach your CRM. Dedupely doesn't block records from being created. It finds and merges new duplicates on a schedule you choose, with Auto Merge for searches you've tested and Auto Match for the ones that need a person.
Where do new duplicates come from?
| Source | Why it creates duplicates |
|---|---|
| Web forms | People use a different email, a nickname, or a typo when they fill in a form again. |
| List imports | Rows without an email, domain or record ID create new records instead of updating existing ones. |
| Integrations and syncs | Some CRM checks don't cover records created through the API, like HubSpot's company domain check. Two-way syncs can re-create records after a merge. |
| Call and dialer tools | They often push just a phone number, with a blank or placeholder name. |
| Enrichment and prospecting tools | They add the same person with a different email or company spelling. |
| Reps creating records by hand | They don't search first, or they type the name differently. |
To find your biggest source, look at a few of your current matches in Dedupely and check a source field on each record (like Original Source, Lead Source or Created By). Add it to Visible Fields so you can see it in your results.
What does your CRM already catch?
Keep your CRM's own checks on. They stop some duplicates before they exist, but each one has gaps:
| CRM | What it checks | What gets through |
|---|---|---|
| HubSpot | Contacts with the same email, and companies with the same domain. You can also make up to 10 custom properties unique per object. | Different emails for the same person; companies created through the API or a sync app; custom unique properties aren't enforced on forms. |
| Salesforce | Duplicate Rules alert or block when a record is created or edited. Imports and API calls that match an existing record are blocked, unless the integration is set to save duplicates anyway. | Rules set to alert, which reps click past; typos and nicknames the matching rule misses; duplicates inside the same import; converted leads (unless Apex Lead Convert is on); records added by Lightning Sync or Einstein Activity Capture. |
| Pipedrive | Flags people with the same name plus a shared phone, email or organization. Imports can merge duplicates on a few set fields. | Thin records from call tools with no name, and spelling differences. |
| Close CRM | No duplicate detection in the app. | Everything that doesn't match exactly. |
| Nimble CRM | Imports merge on email, or on name when no email matches. | Records added after an import. |
| Google Contacts | Suggests duplicates it's confident about. | The pairs it isn't sure about. |
For the full comparison, see CRM Duplicate Tools Compared.
How to fix the sources you control
- Forms: use your CRM's native form or integration, and make email a required field.
- Imports: dedupe lists before you import them, and always map a unique identifier (email, domain or record ID). See Dedupe a List Before You Import It.
- Integrations: check each tool's settings for a "match existing records" option, and find out which field it matches on.
- Reps: ask them to search before creating a record, and agree on how company names are written.
How to merge the duplicates that still get through
Build one Duplicate Search per source, based on the data that source sends, and automate each one based on how it did in your tests:
- Create the search. For example, for a dialer that sends only a phone number: Phone Number (Exact, or Similar to compare the last 7 digits), with a filter so one or more records in the match have a blank First Name. Add Visible Fields like Last Name, Email and Create Date to compare before merging.
- Test it by hand. Merge 1–2 matches, check the result, and reject the false matches.
- Turn on Auto Match if a person needs to check its matches. You're emailed when they reach the number you set.
- Turn on Auto Merge once the search stops finding false matches. Choose a schedule that matches the source:
| Source | Schedule |
|---|---|
| Web forms | As soon as possible: Auto Merge runs in a continuous loop, so new duplicates are merged soon after they reach Dedupely |
| Dialer and call tools | Daily, before your reps start their day |
| Monthly imports | On a set day of the month, just after your usual import |
New CRM records usually reach Dedupely within 1–5 minutes, and Dedupely refreshes your data every 10–30 minutes, so even "As soon as possible" isn't instant. See Auto Merge and Auto Match.
Example setup
An ecommerce team gets duplicate contacts from three places: checkout, a newsletter form and a call tool.
| Source | Duplicate Search | Automation |
|---|---|---|
| Newsletter form | First Name: Exact · Last Name: Exact · Email: Exact | Auto Merge, As soon as possible |
| Checkout (personal and work emails) | First Name: Similar Words (Any Order) with Match nicknames · Last Name: Exact · Phone Number: Similar | Auto Match, notify at 10 matches |
| Call tool | Phone Number: Similar · Filter: First Name is blank, Applies to one or more | Auto Merge, daily at 7 AM |
Merge Rules: Primary Record set to Most Field Data, so the complete record wins over the thin one from the call tool.
Common mistakes
- Only merging, never fixing the source. The same duplicates come back every week.
- Turning off your CRM's own checks. They're free, and they stop the easy duplicates.
- One search for every source. Each source sends different data, so each needs its own fields and match options.
- Auto Merge on an untested search. Start with Auto Match, then switch.
- Forgetting two-way syncs. If an integration syncs records back, check how it handles merges first. See Merging duplicates with an active HubSpot-Salesforce Integration.
FAQs about stopping duplicates
- Can Dedupely block a duplicate from being created? No. Dedupely finds and merges duplicates after they reach your CRM. With Auto Merge on the "As soon as possible" schedule, they're merged soon after they arrive.
- How quickly does Dedupely see new records? Usually within 1–5 minutes. Data refreshes every 10–30 minutes.
- Should I turn off my CRM's own duplicate checks? No. Keep them on, and use Dedupely for what they miss.
- Is Auto Merge on the free trial? Yes. It uses your 20 free merges and stops when they run out. Auto Match is included too.
- Can I see which source creates the most duplicates? Add a source field like Original Source to Visible Fields and compare your matches.
Next up
Auto Merge Merge new duplicates in the background on a schedule, once you've tested your Duplicate Search.Still stuck?
Our team helps on every plan and during the free trial, by chat, email or Zoom.
- Need help? Talk to support
- Questions about plans? Book a Zoom with sales
- Reading this with a model? Read this page as Markdown
- Feeding it the whole site? Read llms.txt
Or sign up with Google
HubSpot Marketplace 
