Data 360 Tools · Step 1 of 2
Data harmonization, visualized.
One customer, four systems, four different shapes. Harmonization in Salesforce Data 360 (formerly Data Cloud) maps them onto one model so everything downstream can see a single person. Flip a step off and watch what breaks.
What everything downstream sees
Identity resolution here uses the Balanced ruleset from the simulator: fuzzy first name plus last name, confirmed by an email, phone, or address. Records that can't satisfy it stay separate, even when an email matches.
Sources to the data model
Raw data first lands as-is in Data Lake Objects. Mapping then moves each field onto the shared Data Model Objects on the right. Click any field to trace where it lands.
Next: identity resolution
Harmonization gets every record into the same shape. It doesn't decide who is who. That's the job of identity resolution: match rules compare the harmonized names, emails, phones, and addresses, then link the records that belong to the same person into one unified profile.
The two steps depend on each other. Rules can only match on what was mapped, so an unsplit name or an unmapped phone quietly takes a matching option off the table. And even perfect mapping doesn't help if the rules are too strict to connect Bob with Robert, or loose enough to merge a husband and wife.
Follow Robert into the identity resolution simulator →
Documents need their own kind of preparation before an agent can use them. See vector search, visualized →
All data is fictional. An independent teaching tool, not affiliated with Salesforce; the model is simplified for learning. More Data 360 tools