Salesforce Duplicate Management: Rules, Matching & Merge Guide

March 2026 · 14 min read

A sales rep spends 20 minutes researching an account, crafting a personalized email, and hitting send. Thirty seconds later, a colleague on the same team pings them: "Hey, that's my account." They check Salesforce. Two Account records for the same company, one created by marketing from a webinar list, the other entered manually by sales. Different owners, different activity histories, different Contact records attached to each. Neither one has the full picture.

This happens constantly. And it's not a discipline problem. It's a Salesforce duplicate management problem. Without the right rules configured, duplicates enter your CRM faster than your team can find and merge them.

Salesforce has built-in tools for this. Matching rules, duplicate rules, and a native merge feature. Most orgs either don't use them or use them so loosely that they might as well not be there. This guide covers how to actually set them up, what to do with the duplicates you already have, and how to keep them from coming back.

How Salesforce Duplicate Detection Works

Salesforce's duplicate management system has two layers, and you need to understand both to configure it correctly.

Matching Rules

A matching rule tells Salesforce how to compare records and determine if they're potential duplicates. It specifies which fields to compare, what matching method to use for each field, and how to combine those field-level comparisons into an overall match score.

Salesforce includes standard matching rules for Accounts, Contacts, and Leads. They're decent starting points but usually need customization.

Standard Account matching rule: Compares Account Name and Billing Street using fuzzy matching. This catches "Acme Corp" vs "Acme Corporation" but misses "Acme Inc." vs "ACME" because the fuzzy threshold isn't aggressive enough by default.

Standard Contact matching rule: Compares First Name, Last Name, Email, and Phone. Email is the strongest signal here. If two contacts have the same email address, they're almost certainly the same person.

Standard Lead matching rule: Compares Company, First Name, Last Name, Email, and Title. Similar logic to Contacts but applied to the Lead object.

Duplicate Rules

A duplicate rule defines what happens when a matching rule finds a potential duplicate. You have three action options:

• Allow (with alert): The user sees a warning that a potential duplicate exists but can still save the record. Most orgs start here.
• Block: The user cannot save the record if a match is found. Stricter, but prevents duplicates more effectively.
• Report: Salesforce silently logs the potential duplicate in a Duplicate Record Set for later review. No user-facing alert.

You can also control which channels the rule applies to: user interface, API, or both. This matters because a lot of duplicates enter through integrations and imports that bypass the UI entirely.

Setting Up Matching Rules

Go to Setup, search for "Matching Rules," and either edit the standard rules or create custom ones. Custom rules give you more control over which fields to compare and how.

Recommended matching rules for Contacts

Field	Matching Method	Why
Email	Exact	Strongest dedup signal. Same email = same person, almost always.
First Name	Fuzzy: First Name	Handles "Bob" vs "Robert," "Mike" vs "Michael"
Last Name	Exact	Last names rarely have legitimate variations
Account Name	Fuzzy: Company Name	Catches contacts at the same company with different account spellings

Create two matching rules for Contacts: one that matches on Email alone (high confidence), and one that matches on First Name + Last Name + Account Name (medium confidence). This lets you set different actions for each confidence level.

Recommended matching rules for Accounts

Field	Matching Method	Why
Account Name	Fuzzy: Company Name	Core comparison. Handles suffix and abbreviation variations.
Website	Exact	If two accounts share a domain, they're likely the same company
Billing City	Exact	Disambiguates franchises and companies with generic names

Account matching is trickier than Contact matching. Company names have endless variations. Adding Website as a matching field dramatically improves accuracy because domains are unique. If you have a normalized company name field, use that instead of the raw Account Name for even better results.

Recommended matching rules for Leads

Mirror your Contact matching rules on the Lead object. Same email match, same name + company match. This catches duplicates both within Leads and across Leads and Contacts (cross-object matching).

Cross-object matching is particularly important. A new Lead comes in from a webinar, but that person already exists as a Contact on an Account. Without Lead-to-Contact matching, nobody knows. The Lead gets assigned to an SDR while the existing Account Owner has no idea their contact just raised their hand.

Configuring Duplicate Rules

Once your matching rules are active, create duplicate rules that reference them.

Recommended configuration

1. Create a "Block" rule for exact email matches. If a Contact or Lead has the same email as an existing record, block the save. There's almost no legitimate reason to have two records with the same email address.
2. Create an "Alert" rule for fuzzy name matches. Fuzzy matching catches more duplicates but also produces false positives. Let users override these with a click, but make sure they see the warning.
3. Enable rules for API and bulk imports. Under the duplicate rule settings, check the boxes for when records are created via API. Most orgs leave this off by default, which means every integration and data import bypasses duplicate checking entirely.
4. Create Duplicate Record Sets. Even when you allow duplicates through, configure the rule to create a Duplicate Record Set. This gives you a queue to review later.

Merging Records: Object by Object

Finding duplicates is half the battle. Merging them without losing data is the other half.

Account to Account Merge

Account merges are the most complex because Accounts have the most related records: Contacts, Opportunities, Cases, Activities, Notes, Attachments, and custom child objects.

How to merge Accounts natively:

1. Go to the Account you want to keep (the "master")
2. Click "Merge Accounts" from the action menu
3. Search for and select the duplicate(s), up to 3 total
4. Choose which field values to keep from each record
5. Click Merge

What happens to related records: All Contacts, Opportunities, Cases, and Activities from the merged records move to the master Account. The duplicate Account records go to the Recycle Bin.

Watch out for: Account Team members, sharing rules, and custom lookup relationships. These don't always merge cleanly. Check your master Account after merging to make sure nothing was lost.

Contact to Contact Merge

Contact merges are simpler but still need care.

1. Navigate to the Account that contains the duplicate Contacts
2. Select "Merge Contacts" from the related list or use the Duplicate Record Set
3. Choose the master record
4. Select field values from each record
5. Merge

Activities from all merged Contacts transfer to the master. Campaign Member records are preserved. If both Contacts were in the same Campaign, the earliest membership date is kept.

Lead to Lead Merge

Lead merges work the same way as Contact merges. Go to the Leads tab, find the duplicates, choose a master, select field values, merge.

One thing to watch: Lead conversion history. If one Lead was previously converted and then re-created (it happens), merging the duplicate into it may produce unexpected results. Always check the Lead History after merging.

Handling Duplicates at Scale

Salesforce's native merge works fine for small numbers. If you've got 50 duplicate pairs, you can knock those out manually in an afternoon. But if your data quality audit revealed 5,000 potential duplicate sets, you need a different approach.

Third-party tools

Tool	Strengths	Pricing
Validity DemandTools	Mature, powerful merge rules, bulk operations, good for large orgs	Subscription, contact vendor
Cloudingo	AppExchange native, good UI, automated scheduling	Starting ~$1,000/year
DataGroomr	AI-powered matching, good for complex duplicates	Starting ~$99/month
Duplicate Check	Real-time and batch, cross-object matching	Free tier available

These tools all do roughly the same thing: identify duplicate sets with configurable matching logic, let you preview merges before executing, and process merges in bulk. The differences are in UI quality, matching algorithm sophistication, and price.

Bulk merge process

1. Export your duplicate sets. Use your tool to identify all potential matches and export them to a spreadsheet for review.
2. Review a sample. Check the first 100 pairs manually to make sure the matching logic is right. False positives are expensive to undo.
3. Define merge rules. Which record survives? Usually the one with the most recent activity, or the one with more complete data. Define this as a rule, not a case-by-case decision.
4. Run a test batch. Merge 50-100 pairs and verify the results. Check that related records transferred, field values are correct, and nothing disappeared.
5. Process the rest. Once you're confident in the rules, run the full batch. Most tools process 500-1,000 merges per hour.
6. Verify. Spot-check 20-30 merged records to confirm everything looks right.

Common Duplicate Management Mistakes

Setting rules to "Alert" when you mean "Block." Users click through alerts. Every time. If you want to actually prevent duplicates, block the save for high-confidence matches. Save alerts for lower-confidence fuzzy matches where human judgment is needed.

Not enabling rules for API and imports. This is the single most common oversight. Your web forms, integrations, and CSV imports are probably your biggest source of duplicates, and they bypass UI-based duplicate rules by default. Enable API enforcement.

Merging without a master record strategy. If your team doesn't know which record to keep as the master, they'll make inconsistent choices. Some keep the older record, some keep the newer one, some pick randomly. Define a policy: master record is the one with the most recent Activity, or the one owned by the Account Owner, or the one with the most complete data. Pick one and stick with it.

Ignoring cross-object duplicates. A person can exist as both a Lead and a Contact in Salesforce simultaneously. These are technically on different objects, but they're still duplicates. Use cross-object matching rules to catch Lead-to-Contact matches.

Treating deduplication as a one-time project. You cleaned up 3,000 duplicates last quarter. Great. But without prevention rules in place, you'll have 1,000 new duplicates by next quarter. Deduplication without prevention is an expensive treadmill.

When to DIY vs. Outsource

If you have fewer than 500 duplicate pairs, you can handle this with Salesforce's native tools in a few days. Standard matching rules, a duplicate rule set to Block, and manual merging.

Between 500 and 5,000 duplicates, a third-party tool pays for itself. The time savings from bulk merging alone justifies the cost.

Above 5,000 duplicates, or if your matching logic needs to be complex (handling subsidiaries, international entities, multiple name variations), it's worth bringing in help. We do this kind of cleanup regularly. The tricky part isn't the merging itself, it's building the matching logic that catches real duplicates without creating false positives that merge records that shouldn't be merged.

Common Questions About Salesforce Duplicate Management

Related: All Salesforce Guides | Duplicate Contacts Guide | Company Name Normalization | Data Quality Audit