Identity Resolution in Salesforce Data 360: The Keychain Behind the Unified Profile

Data 360 (formerly Data Cloud) has existed for several years, and most professionals now recognize why identity resolution matters. Customers appear across multiple systems: CRM, marketing automation, service desks, loyalty programs. Each system maintains its own version of the data. Identity Resolution connects those records into a comprehensive customer view.

But understanding the mechanics is equally important. Beyond the conceptual “why,” the technical implementation determines whether your unified profile stays trustworthy or becomes another source of confusion.

The Keychain Analogy

Consider your personal keychain. You carry keys for your house, your car, your gym locker, maybe a storage unit. Each key has a unique cut. You manage them all because they attach to a single keychain.

Data 360’s Identity Resolution works the same way:

• Keys are individual records from different systems: a CRM contact, a marketing subscriber, a loyalty ID
• The keychain is the resolved identity, linking all of those records under one profile

The keychain doesn’t reshape the keys. It doesn’t merge your car and house keys into one super-key. Each location retains its own contents. The keychain simply ensures you can access all of them with confidence, knowing they belong to the same person.

This is a sharp contrast to the Golden Record approach. Golden Record methodology merges data into a single master record: one name, one address, one phone number. That creates trade-offs. You have to pick “the truth” for every field, even when multiple values have legitimate reasons to coexist.

Data 360 links records rather than collapsing them. Jane may have two email addresses and two physical addresses. You know both belong to the same Jane, and you retain the ability to use both. Her billing address and her shipping address are different for a reason.

Matching on Common Attributes vs. PartyIdentification

Most people are familiar with the basic form of Identity Resolution: matching on everyday attributes like Name + Email, Name + Address, Name + Phone. These rules are intuitive. If two records share those combinations, they probably belong to the same person.

But what happens when a source system doesn’t store name, email, phone, or address? What if it only contains identifiers, some of which may be shared across multiple systems?

This is where PartyIdentification becomes critical.

PartyIdentification is structured storage for specialized identifiers:

• Loyalty ID
• Driver’s License Number
• Passport Number
• External CRM or ERP account numbers

These identifiers might not be the primary key in their source system, but they provide strong identity proof. A customer might appear with different contact details in two systems but carry the same loyalty membership number, and that number confidently links them.

PartyIdentification gives Data 360 a structured way to store and classify these identifiers alongside, or instead of, common attribute matching.

Why This Architecture Matters

• Context is preserved: Data remains intact without forced merging
• Complexity is handled: Customers switch systems, IDs, and contact methods. The keychain model accommodates that
• Downstream use cases work: Segmentation, activation, personalization, and AI all depend on knowing customer identity with confidence

Identity Resolution in Data 360 relies on deliberate linking, not hidden algorithms. The keychain model isn’t just a useful analogy. It’s the design philosophy behind how unified profiles are built and why they can be trusted.

What that means in practice, from which email a customer actually receives to why a bad match is easy to undo, is in Data 360 doesn’t build a golden record. So what?

Before you configure a single rule, though, the prep work matters more than the model. I wrote about that in planning for identity resolution in Data 360.

For a lighter take, I ran into the exact same problem in a shoebox of family records.

To see the keychain in action, try the identity resolution simulator. It shows every source record, its Unified Link, and the Unified Individual it lands on as you change the rules. For the step before it, when each source’s records get mapped onto one shape, see the data harmonization visualizer.