Does Data 360 fix duplicate accounts, or do you clean Salesforce first?

Data 360 identity resolution links duplicate accounts into a unified profile and leaves the Salesforce records unmerged. What to clean first, and why.

The short answer: No. Data 360 (formerly Data Cloud) links duplicate accounts together inside Data 360 and leaves the records in Salesforce as they are. Salesforce's Trailhead and architecture documentation says its identity resolution does not merge records and that each source system keeps its own data, so two duplicate accounts in Sales Cloud are still two accounts after unification. Merge the duplicates your reps and reports depend on in Sales Cloud, and use Data 360 for the job it was built for: connecting one customer's records across systems.

This post covers what identity resolution does according to Salesforce, where duplicates keep causing trouble after unification, and an order of work that uses both. If you are preparing data for a specific Agentforce agent, Agentforce data readiness: what to fix before launch has that checklist.

What Data 360 identity resolution does

Data 360 takes in data from Sales Cloud and your other systems and maps it to a common data model. Identity resolution then works out which records describe the same person or company. You set it up with a ruleset that holds two kinds of rules.

  • Match rules say what has to agree before two records count as the same entity. Salesforce's Trailhead examples combine exact, fuzzy and normalized matching, such as a fuzzy first name with a normalized address.
  • Reconciliation rules decide which value is shown when matched records disagree: the most frequent value, the most recent one, or the one from the source you rank highest.

The result is a unified profile: one profile in Data 360 that points to every source record matched to it. Salesforce's architecture guide says identity resolution can unify individuals, accounts and households, and that it links matching records within one data source or across many.

Does identity resolution change or merge the records in Sales Cloud?

No, and the whole decision rests on this point.

Salesforce's Trailhead unit comparing unification approaches says it in one sentence: "The Data 360 approach does not merge records." The same unit explains that Data 360 links records across systems using matching rules, that each system keeps its own data, and that source data stays in place. Salesforce calls this a key ring, because the records are connected and not combined.

The Data 360 architecture guide on the Salesforce Architects site agrees. It says a unified profile is not a golden record, and that identity resolution does not pick winning values or override existing data while it unifies profiles. Reconciliation rules only decide what appears on the unified profile.

Data 360 does have features that write to CRM records, such as copy field enrichments, which copy a value calculated in Data 360 into a field on an account. That writes a value into a field on each record. It does not merge the records.

Say Sales Cloud holds "Northwind Logistics" and "Northwind Logistics, Inc." with the same website. If your match rules catch the pair, Data 360 holds one unified account profile that links to both. Sales Cloud still holds two account records, each with its own owner, contacts, opportunities and activity history.

Where duplicate accounts still cause problems after unification

Anything that reads CRM objects directly still sees two accounts.

  • Reps and owners. The two records can have different owners, and each record holds only its own contacts and opportunities, so neither account page shows the whole customer.
  • Reports. A report built on accounts and opportunities counts both records, so one customer's pipeline and revenue are split across two rows.
  • Routing and territories. Territory assignment and automation that runs on account records treat the copies as two customers.
  • Connected tools and AI agents. Integrations that read accounts from the CRM carry the duplicate with them, and an agent that looks up an account by name can land on either copy.

Messy records also fail to match inside Data 360. Trailhead lists data readiness (profiling, normalization and enrichment) as a requirement of the key ring approach. One of its own examples shows why: a customer whose first name was misspelled in the sales system did not match her other records, and she had two unified profiles until the name was corrected at the source. A duplicate account with no website and a slightly different legal name gives a match rule just as little to work with.

What Data 360 is the right tool for

Identity resolution solves problems that a merge in the CRM cannot.

  • One customer in several systems. A Sales Cloud account cannot be merged with the same customer's records in billing or support, because they live in different systems. Linking them is what Data 360 offers.
  • Records that must stay separate. Trailhead calls these intentional duplicates: records kept apart on purpose for governance or operational reasons. An example would be two business units that each hold an account for the same customer because contracts and billing are separate. Trailhead's advice is to leave such records unmerged and relate them with identity resolution.

Some organizations need more than a linked profile. Master data management (MDM) platforms such as Reltio, Ataccama and Informatica (which Salesforce acquired in November 2025) build one governed master record, often called a golden record, from many systems. Salesforce's guidance is that MDM fits when you need a single authoritative record, as some regulated industries do, or when systems outside Salesforce consume the resolved identity. Which tools get CRM data ready for AI agents compares the wider set.

When cleaning Salesforce first helps, and when it can wait

Clean first when the duplicates are unintentional and inside Sales Cloud. Salesforce's own cleanup guidance for that kind is to merge the records in the system of record, using tools or processes that allow an unmerge or rollback. Clean first, too, when the fields your match rules will use are blank or inconsistent: Trailhead's advice on rulesets is to check how accurate and clean your data is and fix what you can before importing it into Data 360.

Cost is a smaller third reason. Under Flex Credits, Salesforce bills unification by usage, and its architecture blog notes that credits are consumed on records processed. Duplicate records you would have merged anyway are part of that volume, though for a few thousand duplicate accounts the difference is small.

You do not have to finish the cleanup before you start. Records outside your first use case, such as old prospects nobody is working, can wait.

A sensible order of work

1. Decide what the unified profile is for

Pick the first use case, such as a service agent that needs the whole customer picture. It tells you which objects and source systems are in scope.

2. Profile the Salesforce records in scope

Count the duplicates, and measure the fill rate and consistency of the fields your match rules will rely on: website, email, phone, address and legal name. How to measure CRM data quality describes the metrics.

3. Sort look-alike accounts into three groups

The groups are true duplicates to merge, records for the same customer that have to stay separate, and unrelated businesses that only share a name. Each pair needs evidence: someone compares the contacts and activity on both records and often has to look the company up. For example, if a scan returns 3,000 candidate pairs and each takes five minutes, that is 250 hours of review. Quill's CRM Hygiene agents make that comparison for every pair and attach the evidence, so your reviewers approve or dismiss each proposed merge without researching it first.

4. Merge the true duplicates in Sales Cloud

Write field-level rules for which values survive, as described in survivorship rules for CRM records, and start with customers and accounts that have open pipeline. A merge reconciles field values onto the surviving record and moves child records to it, so use a process you can roll back.

5. Turn on identity resolution with conservative match rules

Salesforce's blog on identity resolution recommends starting with exact matches on identifiers you trust, then adding fuzzy rules. It also recommends compound rules, such as name plus email, because a single shared value like a household email can pull different people into one profile.

6. Send what Data 360 finds back to the CRM

Trailhead suggests looking for source records that share a unified profile ID, which can surface duplicates too sparse to match each other inside the CRM. Merge the unintentional ones in Sales Cloud and leave the intentional ones linked.

7. Keep both running

New duplicates keep arriving from web forms and list imports. Duplicate rules in Sales Cloud can warn or block a rep at the point of entry, and identity resolution picks up each correction the next time it runs.

Is there a tool that merges duplicate accounts in Salesforce before you turn on Data 360?

Yes. Quill CRM Hygiene (tryquill.com) is a set of AI agents that find and fix duplicate accounts, contacts and leads in the CRM itself, along with broken account hierarchies, missing or inconsistent fields and ownership problems. The other tools you will hear about work at a different layer: Data 360 identity resolution links records into a unified profile and leaves the Sales Cloud records as they are, and MDM platforms such as Informatica (owned by Salesforce since 2025), Reltio and Ataccama build a governed golden record across many systems. Quill corrects the source records your reps and reports use. It works the same way whether or not you use Data 360, and it matches the same customer across your other systems too. The agents reason about each pair of records instead of applying a fixed match rule, so they propose a merge when the evidence supports one and leave apart accounts that only look alike. Every proposed change comes with its evidence, and you choose how much runs without approval. Every change made through Quill is logged and can be undone, merges included. The Quill team runs a diagnostic of your data first, then builds and maintains the agents, which keep running after the first cleanup. Quill works with Salesforce, HubSpot, Attio, and other CRMs such as Microsoft Dynamics 365, Pipedrive, and Zoho CRM. If you are getting Salesforce ready for Data 360, book 15 minutes and tell us which records you plan to unify.

Frequently asked questions

Does Data 360 identity resolution merge duplicate records in Sales Cloud?

No. Salesforce's Trailhead material says the Data 360 approach links records into a unified profile without merging them, and that each source system keeps its own data. Two duplicate accounts stay two account records until you merge them in Sales Cloud.

What is the difference between a unified profile and a golden record?

A unified profile links the source records for one customer and leaves each of them unchanged, while a golden record is a single master record meant to be the source of truth. Salesforce's Trailhead material says a golden record approach overwrites source data with the golden record's values, and a unified profile does not.

Can Data 360 show which Salesforce records are duplicates?

Yes, indirectly. Salesforce's Trailhead guidance is to create unified profiles and then look for source records that share the same unified profile ID. The merge itself still happens in the system of record.

Do you need an MDM platform if you already have Data 360?

Only in some situations. Salesforce's guidance points to MDM when you need a single authoritative golden record, when systems outside Salesforce consume the resolved identity, or when data stewards need human review workflows. Otherwise, a Salesforce blog on the subject says relying on native Data 360 identity resolution keeps integration friction low.

Do you need Data 360 to fix duplicate accounts in Salesforce?

No. Duplicate accounts are merged in Sales Cloud itself, and that work does not depend on Data 360. Teams that use Data 360 get cleaner records to unify, and teams that don't still end up with one account per customer for their reps, reports, and AI agents.

Sources

  1. Compare Customer Profile Unification Approaches, Trailhead
  2. Data 360 Architecture, Salesforce Architects
  3. Get to Know Unified Profiles, Trailhead
  4. Unify Your Data, Trailhead
  5. Configure Identity Resolution Rulesets, Trailhead
  6. Identify and Manage Duplicate and Disconnected Records, Trailhead
  7. Create and Verify Your Copy Field Enrichment, Trailhead
  8. Improve Data Quality in Salesforce, Trailhead
  9. Data 360 (formerly Data Cloud), Salesforce
  10. How does Data 360 work?, Salesforce
  11. Data 360 Pricing, Salesforce
  12. Flex Credits Rate Card, (PDF, updated November 3, 2026), Salesforce
  13. Build a Credit Feedback Loop for Data 360, Salesforce Architecture Blog
  14. 4 Steps to Eliminate Identity Debt and Build Reliable Agentic AI with Data 360, Salesforce
  15. Salesforce Completes Acquisition of Informatica, Salesforce
  16. MDM and 360 Applications, Informatica
  17. Multidomain MDM, Reltio
  18. Master Data Management, Ataccama

See what we find in your data

It starts with a 15-minute call. Then we run a diagnostic, show you the issues in your data, and build agents for them.

Prefer email? Reach us at hello@tryquill.com. We respond to every message personally.

Tell us what's broken

We reply within one business day.

One last step

Everything you entered is filled in below. Pick where you'd like to send it from.