Salesforce Data Cloud (Data 360): the enterprise guide for 2026

Salesforce Data Cloud, renamed Data 360 at Dreamforce 2025 (October 2025), is the native data engine of the Salesforce Platform. It brings together information from any source (structured and unstructured), builds a unified profile of every customer in real time, and delivers this context to Salesforce clouds, external systems, and, above all, Agentforce AI agents. In one sentence: it is the layer that turns scattered data into action within the CRM.

For enterprise operations that comprise dozens of systems and millions of records, this difference separates a CRM that "stores data" from one that acts on it. In this guide, you will find out what Data Cloud is in 2026, how it works internally, what Zero Copy is, why it became a prerequisite for Agentforce, and how a serious implementation begins.

What is Salesforce Data Cloud (Data 360)?

Data Cloud is Salesforce's native customer data platform (CDP). Unlike a third-party CDP, it is built within the Salesforce Platform itself, uses the same CRM metadata model, and needs no external connector to communicate with Sales Cloud, Service Cloud, or Marketing Cloud. It ingests data from a variety of sources, resolves identity (bringing together the 'John Smith' that appears in five systems into a single profile), and makes that unified profile available to any platform application.

Three functions define the product:

• Unify: consolidate CRM, ERP, e-commerce, customer-service, and digital-channel data into a single source of truth.
• Understand: apply segmentation, calculated insights, and AI to this data.
• Activate: send the result back to where action happens, whether in a campaign, customer service interaction, or an Agentforce agent.

Data Cloud or Data 360? What changed with the name

At Dreamforce 2025, Salesforce began calling Data Cloud Data 360, under the Agentforce 360 umbrella. The capabilities remain the same; what changes is the name and positioning, now explicitly as the data layer that supports AI agents. The product has had several names over the years (Customer 360 Audiences, Salesforce CDP, Genie, Data Cloud), so do not be surprised to find Data Cloud and Data 360 used together in documentation and searches. Throughout this text, we use 'Data Cloud,' which is still the term most searched in the Brazilian market.

How Data Cloud differs from a CRM and a data warehouse

The confusion is common, and understanding the role of each component prevents the wrong purchase. CRM is where work happens, the data warehouse is where historical data is analysed, and Data Cloud is the bridge that activates real-time data in the customer experience.

Layer	Purpose	Optimized for
CRM (Sales, Service)	Run processes: sales, customer service, cases	Day-to-day transactions
Data warehouse (Snowflake, BigQuery)	Store and analyze large volumes of historical data	Analytics and BI
Data Cloud (Data 360)	Unify and activate data in real time	Identity resolution and activation in the experience

How Data Cloud works, in five steps

Behind the marketing, Data Cloud follows a predictable flow. Understanding these stages helps you estimate the effort required and engage on equal footing with any vendor.

1. Ingestion. Connectors and data streams bring in data from Salesforce, legacy systems, data lakes, and external sources. Ingestion can be event-based (real time) or batch-based.
2. Modeling. Raw data arrives as Data Lake Objects (DLOs) and is mapped to the Customer 360 model, becoming standardized Data Model Objects (DMOs).
3. Identity resolution. Deterministic and probabilistic rules combine records from the same customer into a unified profile, eliminating duplicates.
4. Insights and segmentation. With the profile ready, you create segments and Calculated Insights (pre-calculated metrics, such as customer lifetime value).
5. Activation. Processed data is sent back to Salesforce clouds, marketing channels, automations, and increasingly to Agentforce agents.

The advantage for organisations already running Salesforce is that this flow takes place within the platform, without having to build and maintain expensive external data pipelines.

Zero Copy: data without moving it, without ETL

O Zero Copy (or Zero ETL) is one of Data Cloud's most relevant and least explained features in Portuguese. The idea is to access and query data that resides in an external data warehouse without copying, moving, or duplicating nothing. Data remains at the source, and Data Cloud creates virtual tables that point to it.

In practice, there are two paths:

• Data Federation (data ingestion): Data Cloud queries the data in the external warehouse without bringing it inside.
• Data Sharing (data egress): the external warehouse accesses Data Cloud objects (DMOs, Calculated Insights) through virtual tables, without copying CRM data.

Salesforce positions itself as a Zero Copy pioneer ('first to market,' in the company's own terms), with native support for Snowflake, Databricks, Google BigQuery, and Amazon Redshift, as well as a partner network that expands the ecosystem. For an enterprise company, this means less data copying, less latency, and lower data-movement costs.

Approach	How data moves	Cost and risk
Traditional ETL	Extracts, transforms, and copies to another location	More duplicate storage, more latency, more maintenance
Zero Copy	Data remains at the source; real-time federated query	Less copying, less lag, simpler governance

A critical perspective is essential: Zero Copy shines when there is a consolidated corporate data warehouse. When data is scattered across spreadsheets and systems without governance, the benefit only appears after getting the house in order—and that is where consulting work comes in.

Why Data Cloud is the foundation of Agentforce

This is the point that defines Salesforce's data strategy in 2026. An AI agent is only as good as the context it can access. Without unified data, Agentforce responds generically or, worse, makes mistakes. With Data Cloud behind it, the agent sees the customer's complete history and acts on a sound basis.

Three capabilities make this possible:

• Data Graphs: pre-calculated, materialized views of the customer data model. They provide low-latency context to Agentforce and Prompt Builder without having to run complex queries on every interaction.
• Unstructured data and vector search: Since 2024, Data Cloud has ingested PDFs, emails, call transcripts, and other documents, applying semantic search to ground generative AI. This matters because, according to market estimates (IDC, Gartner, and others), around 90% of corporate data is unstructured, and a Deloitte survey (2019) indicates that only 18% of companies can leverage it.
• RAG (retrieval-augmented generation): the agent retrieves the right information from Data Cloud when responding, rather than "hallucinating" based on a generic model.

That is why, in enterprise projects, the conversation about Agentforce almost always first runs into a data question: is the foundation unified enough to support AI agents? Data Cloud is Salesforce's answer to that question.

Use cases by department

Data Cloud is not an isolated IT project; it delivers value across every business area:

• Sales: lead qualification with intent signals (campaign engagement, website behaviour) updated in the CRM in real time, feeding the Sales Cloud.
• Service: first-contact resolution because the human agent or AI agent sees the unified history within the Service Cloud.
• Marketing: dynamic segmentation and journeys that respond to recent behavior, integrated with Marketing Cloud.
• Data and BI: a single semantic layer (with Tableau) to standardise metrics and translate technical data into business language.

How much it costs: the credit model

Data Cloud is licensed by consumption, in credits. Instead of paying per user, you pay for what you process: ingestion, identity resolution, segmentation, activation, transformations, and vector search consume credits. In September 2025, Salesforce announced the separation between computing and storage credits, so it is worth modelling both.

A few points that help with planning:

• Higher Salesforce editions already include a initial credit and storage allowance to start using Data Cloud without additional contracting.
• There are specific configurations for larger scenarios, such as Data Cloud One, aimed at architectures with multiple orgs.
• Salesforce offers a official pricing calculator to estimate consumption.

A candid note: Salesforce does not publish a price list in Brazilian reais, and the actual cost depends on data volume, use cases, and negotiation. A fixed figure is only available in a quote from an official partner. The most costly mistake is not the license price; it is sizing credits without a usage plan and discovering consumption only on the invoice. This is, by far, where good consulting pays for itself.

What your Salesforce team needs to know about 2025 and 2026 updates

• Data Cloud became Data 360 (Dreamforce 2025), within Agentforce 360.
• Tableau Semantics: a new semantic layer to standardise metrics and indicators on Data Cloud.
• Unstructured data and vector search already generally available, expanding what AI can interpret.
• New pricing structure (Sep/2025), with storage separated from computing credits.
• Market recognition: Salesforce was ranked as a leader by Forrester, Gartner, and IDC in recent CDP analyses, and reported that recurring Data Cloud and AI revenue surpassed US$900 million by the end of fiscal year 2025.

How to implement Data Cloud (and the role of a consultancy)

Buying Data Cloud is easy. Extracting value from it is a project. Most implementations that stall make one of these mistakes: connecting everything without a clear use case, modelling data weakly, or exhausting credits through consumption that nobody anticipated. A well-run project follows a more measured path.

1. Assessment and use cases. Before connecting any source, define what the business wants to solve (faster customer service, real-time marketing, a foundation for Agentforce) and set priorities.
2. Architecture and modeling. Map sources, design the data model, and decide what comes in through ingestion and what remains through Zero Copy.
3. Identity resolution and governance. Define unification, quality, and privacy rules from the outset—not as a patch.
4. Activation and measurement. Put data to work in campaigns, customer service, and agents, and measure credit consumption so there are no surprises.

How official Salesforce Partner, WeeNow works at each of these stages, from implementation from scratch to go-live and ongoing support. The work of a serious consultancy is not to promise magic; it is to size the project honestly, avoid wasting credits, and ensure Data Cloud supports what comes next, including Agentforce. To understand where Data Cloud fits your scenario, it is worth starting with a conversation about Salesforce operation current.

Frequently asked questions about Salesforce Data Cloud

What is Salesforce Data Cloud?

It is Salesforce's native customer data platform (CDP). It ingests, unifies, and activates data from any source into real-time customer profiles, serving as the foundation for CRM, Agentforce, and the entire Salesforce Platform.

Are Data Cloud and Data 360 the same thing?

Yes. At Dreamforce 2025 (October 2025), Salesforce renamed Data Cloud to Data 360 as part of Agentforce 360. The capabilities remain the same; only the name and positioning changed.

Is Data Cloud mandatory to use Agentforce?

It is not technically mandatory, but Agentforce agents rely on Data Cloud to access the full customer context (unified data, unstructured data, Data Graphs). Without this foundation, the quality of agent responses and actions drops significantly.

What is Zero Copy in Data Cloud?

It is the ability to query data in data warehouses such as Snowflake, Databricks, BigQuery, and Redshift without copying or moving it. Data remains at the source, and Data Cloud creates virtual tables to access it in real time, reducing cost and latency.

What is the difference between Data Cloud and a data warehouse such as Snowflake?

Snowflake stores and processes large volumes of historical data for analysis. Data Cloud is optimised to activate real-time data in the customer experience (sales, service, marketing, AI agents), with native identity resolution. They are complementary, not competitors.

Does Salesforce Data Cloud have a free version?

Not exactly free, but higher Salesforce editions already include an initial allowance of credits and storage to start using Data Cloud without a separate contract. Above that allowance, the model is credit-based consumption.

How much does Salesforce Data Cloud cost?

The model is consumption-based (credits), with storage charged separately since 2025. Actual cost depends on data volume and use cases, and Salesforce does not publish pricing in Brazilian reais. For a fixed amount, you need a quote from an official partner.

How do you start a Data Cloud project?

The recommended path is to define priority use cases, map data sources, and design the architecture before connecting everything. A partner consultancy helps size credits, model the data, and avoid mistakes that blow the budget.