Salesforce Data 360 Explained: What Every Admin Needs to Know Before Enabling It in 2026

Salesforce Data Cloud is now officially called Data 360. The name change reflects what it has become: not a standalone add-on but the data layer that the entire Customer 360 platform — and every Agentforce agent you will ever deploy — runs on. Most orgs enabling it for the first time make the same mistake: they skip the three foundations that determine whether it actually works, and go straight to activation.

What Salesforce Data 360 actually is

Data 360 is Salesforce's unified customer data platform (CDP), built natively inside the Salesforce platform — not a separate system bolted on via API. It ingests data from multiple sources, maps and harmonises it to a shared schema, resolves identities across sources to create a single unified profile per customer, and activates those profiles for segmentation, personalisation, and AI grounding.

The reason it matters now more than ever: Agentforce agents are only as intelligent as the data they can ground in. An Agentforce agent that only sees your CRM records can answer questions about what is in Salesforce. An Agentforce agent that grounds in a unified Data 360 profile can reason about the full customer journey — across marketing, service, commerce, and web behaviour — before it responds. That is the difference between a chatbot and something that actually deflects cases.

The name changed from Data Cloud to Data 360 in 2026 to signal its role as a cross-platform data layer, not a product silo.

Foundation 1: Data ingestion and mapping

Data 360 needs data before it can unify anything. The ingestion layer connects external sources to your org and maps incoming fields to Data 360's canonical data model.

A critical 2026 pricing change: structured data from Salesforce CRM is now ingested for free. If your primary use case is grounding Agentforce in your CRM data — Accounts, Contacts, Cases, Opportunities — you can do that without consuming paid credits. This removes the most common early objection to enabling Data 360.

External data (web analytics, marketing platforms, commerce data, data lake exports) still consumes credits. Production-ready connectors support batch ingestion from CSV and Parquet files, data lakes, and scheduled file drops without custom ETL work. The mapping step — where you define which incoming fields correspond to which Data 360 schema fields — is where precision matters. Sloppy mapping here creates data quality problems that compound through every downstream step.

Foundation 2: Identity resolution

Identity resolution is the process that links data about the same person across multiple sources into a single unified profile. It is, without question, the most technically consequential configuration decision in a Data 360 implementation — and the one most frequently skipped or under-built.

You configure identity resolution through a ruleset: a set of match rules and reconciliation rules. Match rules define how Data 360 identifies that two records from different sources belong to the same person. A typical ruleset uses email exact-match as the highest-priority rule, falls back to phone matching, and uses probabilistic name-and-address matching as a last resort.

Reconciliation rules determine what value gets written to the unified profile when sources disagree. Common strategies: keep the most recent value (good for email address), keep the most frequent value (good for preferred name), or prioritise data from a trusted source — CRM over web data, for instance.

The failure modes at either extreme are predictable and damaging. Under-built rulesets produce duplicates — the same customer split across multiple profiles. Over-built rulesets produce over-merges — different customers incorrectly combined into one profile. Either one breaks segmentation, personalisation, analytics, and ROI reporting. Most Data 360 implementations that fail in production do not fail because a connector stopped working — they fail because identity resolution was not configured with the same rigour as the ingestion layer.

One technical requirement to flag: you must map incoming customer information streams to the Individual ID field before identity resolution will run correctly. This mapping is easy to miss in the rush to get the first segments built.

Foundation 3: Activation

Activation is where unified profiles become useful — sending segments and calculated insights to Marketing Cloud, Service Cloud, Commerce Cloud, and Agentforce. This is the step most teams want to do first. Do not. Activation built on top of an incomplete ingestion layer or an unreliable identity resolution configuration will surface bad data everywhere it goes, simultaneously, at scale.

The correct sequence: plan use cases → provision correctly → connect and map data sources → configure identity resolution → validate unified profiles → build segments and insights → activate.

What Data 360 costs in 2026

Data 360 uses a consumption-based model:

• Processing credits: $500 per 100,000 credits — consumed by ingestion operations (external data), identity resolution runs, segmentation jobs, and activation events
• Storage: $23 per month per terabyte
• Structured Salesforce CRM data: Free to ingest (updated in 2026)
• Data 360 One premium add-on: approximately $60,000 per year for advanced use cases

Mid-market organisations running Data 360 alongside Agentforce typically spend $5,400 to $14,600 per month fully loaded — making the Data Cloud line item larger than the Agentforce licence line item for most deployments.

When you actually need it

You need paid Data 360 credits when your use case requires any of the following: unified profiles that combine CRM data with web, marketing, service, or commerce data; high-scale event processing (millions of events per day); cross-cloud segmentation; calculated metrics built from multiple source fields; or Agentforce agents that need to reason about the full customer journey rather than just CRM records.

You can start without paid credits if: your Agentforce agents only need CRM record context; you are just testing whether Data 360 fits your architecture; and your data sources are all inside Salesforce.

For teams using our Synapse MCP integration package to connect Claude or other external AI tools to Salesforce, Data 360 unified profiles are the grounding layer that makes cross-system AI responses reliable — connecting Agentforce reasoning to data beyond what is directly accessible inside your org.