Agentforce Data Cloud: How They Work Together (2026)

If you've started evaluating Agentforce, you've probably run into the same confusing advice: some people say Data Cloud is mandatory for it to run, others say you don't need it at all.

The reality lies somewhere in the middle.

While Agentforce relies on Data 360 capabilities behind the scenes, you don't necessarily need a full enterprise Data 360 implementation to get started. The right choice depends on your AI use case, your data architecture, and how much customer context your agents need.

In this guide, we'll explain how Agentforce and Data 360 work together, when the free Data 360 allocation is enough, when a paid implementation makes sense, and how RevOps teams should evaluate the architecture before investing.

Data Cloud is now Data 360: A quick context note

In October 2025, Salesforce rebranded Data Cloud as Data 360 as part of its broader Agentforce strategy. The underlying platform didn't fundamentally change—Salesforce simply aligned its customer data platform more closely with its AI vision. As a result, you'll still see both names used throughout Salesforce documentation, Trailhead modules, partner blogs, and community discussions.

We use "Data 360" throughout this article since that is the current product name.

What is Salesforce Data 360?

At its core, Salesforce Data 360 is Salesforce's unified customer data platform (CDP). Its job is to bring together customer information from Salesforce CRM, marketing systems, product usage data, external warehouses, documents, and other business applications into a single, trusted customer profile.

This directly affects your Agentforce deployment. AI agents are only as reliable as the data they can access. With Data 360, Agentforce can retrieve information from a unified data layer that represents a customer's complete relationship with your business, rather than reasoning over disconnected customer records.

These four Data 360 capabilities make it important for Agentforce implementation:

Unified data ingestion

Data 360 connects structured and unstructured information from sources including Salesforce applications, external databases, cloud data warehouses such as Snowflake, marketing platforms, and business applications. Instead of copying every dataset into Salesforce, organizations can also use zero-copy data federation to query data where it already lives.

Identity resolution

One customer often exists as multiple records across CRM, marketing automation, billing, support, and product systems.

Data 360 uses identity resolution to match and merge these fragmented records into a single customer profile. This gives Agentforce a much more complete understanding of who it's interacting with instead of relying on whichever record happens to be queried first.

Vector search for unstructured content

80% of the world’s data is unstructured, while fewer than one in every five organizations use it to their advantage. For RevOps teams, this reality creates an operational blind spot: AI agents may know who the customer is from CRM records, but not why a deal is blocked, what was promised on the last sales call, or what was agreed in a contract.

Data 360 also ingests and indexes this unstructured data: knowledge articles, PDFs, contracts, support documentation, emails, call transcripts, and more, so Agentforce can enable more accurate responses and better downstream decisions.

Real-time access across your data ecosystem

With Data 360’s zero-copy architecture, you can query data sitting in Snowflake, Databricks, or BigQuery without duplicating it into Data 360, which reduces storage cost and helps you maintain a single source of truth at the same time.

How Agentforce and Data 360 work together

Agentforce and Data 360 are designed to solve different problems, but they're most powerful when used together.

At a high level, the architecture looks like this:

1. Revenue data from Salesforce and external systems is connected to Data 360
2. Data 360 creates unified customer profiles through identity resolution and makes structured and unstructured business data available for retrieval
3. When a user asks Agentforce to complete a task, the Atlas Reasoning Engine determines what information it needs and retrieves the relevant context
4. Atlas plans the appropriate sequence of actions and invokes the required tools or workflows
5. Throughout the interaction, the Einstein Trust Layer applies security, governance, and grounding controls to help ensure responses are based on trusted enterprise data

Step 1: Data 360 brings customer data together

Most organizations don't keep customer data in one place.

Sales data lives in CRM. Marketing engagement sits in automation platforms. Product usage may live in Snowflake or Databricks. Support teams rely on knowledge articles and case histories, while finance maintains billing information elsewhere.

Data 360 connects Salesforce and external sources into a unified customer data layer, and depending on the use case, it can enrich Salesforce org data back into CRM records and pages.

Step 2: Identity resolution creates a trusted customer profile

The same customer often appears multiple times across enterprise systems.

A person might exist as a Lead in Salesforce, a Contact in Marketing Cloud, a billing customer in an ERP, and a user in a product database. Without connecting those identities, an AI agent only sees fragments of the relationship.

Data 360's identity resolution can match and unify these records into persistent customer profiles, reducing duplicate identities and giving Agentforce a more complete understanding of the customer before it responds.

Step 3: Atlas retrieves the right business context

Once a request reaches Agentforce, the Atlas Reasoning Engine that powers the Agents first determines what information it needs to complete the task. Atlas can invoke retrievers and tools to access relevant structured CRM records as well as unstructured content, before reasoning about the next action.

This retrieval-augmented generation (RAG) approach helps ground responses in your organization's current business data rather than relying solely on the model's pretrained knowledge.

🎯 For example, if a customer asks: "Why hasn't my renewal been processed?"

Agentforce can retrieve information such as:

• Current Account and Opportunity records
• Recent support cases
• Contract documentation
• Billing information
• Relevant internal knowledge articles

It then reasons across that context before generating a response or recommending the next action.

Step 4: The Einstein Trust Layer governs every AI interaction

Agentforce is built on Salesforce's Einstein Trust Layer, which provides capabilities such as secure, permission-aware data retrieval, dynamic grounding, data masking, prompt defense, and auditability throughout AI interactions.

These controls help ensure agents access only the data users are permitted to see while reducing the risk of hallucinations and protecting sensitive information.

Do you actually need Data 360 to run Agentforce?

Yes, but not necessarily a full Data 360 implementation.

This is one of the biggest misconceptions surrounding Agentforce. Many sources imply that you need a large-scale Data 360 deployment before you can use AI agents.

In reality, Salesforce includes a free Data 360 allocation through Salesforce Foundations, allowing you to build and deploy Agentforce without immediately purchasing a paid Data 360 tier.

When the free Data 360 allocation is enough

The free allocation works for teams that fit this profile:

• Low-volume agents. A single-agent proof-of-concept, an internal helper for the sales team, or a lightweight customer service bot for a niche use case can run inside this free envelope.
• Salesforce-native data. If the agent only needs data from Sales Cloud, Service Cloud, or other Salesforce clouds, Salesforce-to-Salesforce ingestion is free. You are not burning Data 360 credits on external pipelines.
• Structured CRM queries. Answering "what's the status of case 12345" from existing records does not require the Vector Database or heavy identity resolution.

You still get the Einstein Trust Layer for audit logging and PII masking, which is one of the strongest arguments for using the Salesforce stack in the first place.

When you should consider a paid Data 360 implementation

As AI agents become more deeply embedded in your business, the requirements often extend beyond what the free allocation is designed to support.

A paid Data 360 implementation becomes more compelling when you need:

• RAG over unstructured documents at scale. Vectorizing 10,000 PDFs is not the free-tier use case.
• Identity resolution across non-Salesforce sources. Unifying customer records from a HubSpot instance, a Snowflake warehouse, and Salesforce is where Data 360 credits burn fastest.
• Advanced segmentation and Calculated Insights. Running 100 daily segment refreshes against millions of unified profiles consumes credits at scale.
• Streaming and harmonization of real-time data from warehouses and enterprise applications

At this stage, Data 360 shifts from being an optional enhancement to becoming the underlying data foundation that supports enterprise AI.

Key use cases Agentforce and Data 360 unlock together

While Agentforce can automate tasks on its own, combining it with Data 360 allows AI agents to work with richer customer context, more reliable data, and broader enterprise knowledge. Here are three of the highest-impact use cases for RevOps and customer-facing teams.

Use case #1: Deliver more personalized customer support

Support agents often need information spread across multiple systems before they can answer a customer's question.

With Data 360, Agentforce can retrieve unified customer profiles, previous cases, purchase history, and relevant knowledge articles before generating a response. Instead of switching between applications or escalating simple requests, support teams can resolve issues faster with responses grounded in current customer data rather than generic AI knowledge.

Use case #2: Power enterprise workflows with trusted AI

Many organizations want AI agents to do more than answer questions. They want them to take action.

Because Agentforce operates on trusted business context, organizations can automate workflows such as updating records, recommending follow-up actions, assisting employees, or triggering downstream processes with greater confidence than would be possible using an LLM alone.

The richer the underlying customer data, the more reliable those AI-driven decisions become.

Use case #3: Audit-logged agent execution for regulated industries

For financial services, healthcare, and any org with compliance obligations, the Einstein Trust Layer's per-action logging is one of the strongest reasons to use Agentforce over an open-ended LLM.

Every agent action is traceable, PII is masked before it hits the model, and outputs pass through toxicity filters. Data 360 provides the governed data foundation those audit logs reference back to.

Agentforce and Data 360 shine when the workflow lives inside Salesforce. But what if yours doesn’t? An AI revenue infrastructure platform like Default helps.

Agentforce and Data 360 pricing: What it really costs

Understanding Agentforce pricing can be confusing because you're often paying for two different things at the same time: the AI that performs work and the data platform that supplies trusted context. Depending on how you deploy Agentforce, you may also need to account for Data 360 usage alongside your chosen Agentforce pricing model.

Some tips to simplify this:

• Flex Credits and Data 360 Credits are two separate pools consumed together. Flex Credits fund the agent's actions. Data 360 Credits fund the data operations the agent depends on: ingestion, unification, segmentation, queries. Your Digital Wallet tracks both, but they draw down independently.
• Agentforce usage can consume either Flex Credits or Conversations depending on your usage type and billing setup
• Data 360 has its own pricing models for storing, unifying, and activating customer data, namely Profile-Based Pricing and Flex Credits. Here’s a snapshot:

Hidden costs to watch for

List pricing is only part of the investment when running Data 360 and Agentforce. The larger costs often come from preparing the data and processes that make enterprise AI successful.

• Identity resolution takes planning. Before AI agents can reason across multiple systems, customer identities need to be matched and governed consistently. This is often one of the biggest implementation efforts, not because the technology is difficult, but because business data rarely starts out clean.
• Connecting non-Salesforce systems adds complexity. If your customer data lives across data lakes, ERP systems, or other applications, additional integration work may be required to create a unified data foundation
• Implementation services can rival software costs. Enterprise deployments frequently involve data modeling, governance, security reviews, and implementation partners. These services can become a significant part of your overall investment.
• Knowledge quality matters as much as AI quality. Agentforce can only retrieve information that exists and is well maintained. Many organizations discover they need to clean up knowledge articles, documentation, and CRM records before AI delivers consistent results
• Flex Credits don’t roll over. Unused credits expire when your contract term ends. Over-commit and you leave money on the table; under-commit and you’ll be forced to buy more or deal with overage usage

When Agentforce and Data 360 are the right call

The combination of Agentforce and Data 360 is a strong fit—but it isn't the right architecture for every organization.

It fits when...

• You're already heavily invested in the Salesforce ecosystem
• Customer service, support, or customer experience are primary AI use cases
• You already use—or plan to deploy—Data 360 as your customer data platform
• Your organization has Salesforce administrators or implementation partners who can manage the platform
• Enterprise governance, auditability, and security are key requirements for AI adoption

It may not be the best fit when...

• Your GTM operations span Salesforce, HubSpot, Marketo, Snowflake, and several other systems
• Your priority is operational RevOps workflows such as lead qualification, routing, enrichment, and revenue execution rather than customer service automation via Salesforce
• You need to see value in weeks rather than oversee a multi-quarter implementation
• Budget or internal resources don't support a broader customer data platform rollout
• Your AI agents need to coordinate work across the entire revenue stack rather than primarily within Salesforce

If Agentforce and Data 360 aren't the right fit: An alternative for GTM teams

Agentforce and Data 360 aren’t a strong combination for RevOps teams whose customer and revenue data is spread across multiple CRMs, data warehouses, enrichment providers, product analytics, and dozens of operational tools.

Especially when your AI agents don't just need to answer questions but also qualify leads, coordinate workflows, update systems, and execute across the fragmented revenue stack.

In such cases, a different architecture may make more sense.

Default provides this AI infrastructure for revenue teams.

A unified revenue data layer for GTM

Just as Agentforce relies on Data 360 for trusted customer context, Default provides AI agents with a unified view of revenue operations.

When you connect your CRM to Default, it backfills the records, including people, companies, opportunities, buying signals, and more, into a consolidated data layer and stacks enrichment, scheduling telemetry, routing assignments, and workflow logs on top of it.

If you run Salesforce and HubSpot in parallel, Default reconciles the lead, contact, and campaign member representations into one canonical entity, with field-level source-of-truth control.

The result is a single, live GTM data foundation your agents and workflows read from without SQL, without a data team, and without brittle reverse-ETL pipelines.

AI agents that don't stop at reasoning

Understanding context is only one half of the equation. Once an agent determines what should happen next, it also needs a way to execute that decision. Preferably inside the same systems your RevOps team already operates in.

Default's AI agent, Dot, can orchestrate actions end-to-end across the inbound revenue workflow:

• Lead qualification and disqualification. Dot applies your ICP rules against enriched firmographic, technographic, and intent data. It then either qualifies the lead into a routing workflow or disqualifies with a logged reason.

• Waterfall enrichment. Default runs multi-provider enrichment with fallback logic, so you're not paying for redundant hits or losing coverage when one provider misses.
• Routing decisions with reasoning. Dot assigns leads to the right rep based on territory, account ownership, round-robin balancing, or custom logic, and logs why the routing decision was made, so RevOps can audit every assignment.

• Meeting scheduling and handoff. Dot books qualified meetings directly into rep calendars, respecting availability, buffer rules, and handoff SLAs.

• CRM writes across Salesforce and HubSpot. Dot updates records, creates opportunities, and syncs field-level changes with source-of-truth control. No, you don’t need a reverse-ETL pipeline to do it.
• Workflow triggers and notifications. Dot fires downstream actions in Slack, email sequences, and connected tools when specific conditions are met (e.g., a Tier 1 account signs up, an opportunity stalls, a target persona enters a workflow).

• Answering RevOps questions in natural language. "Why did this lead route to Sarah?" "What happened in this workflow?" Dot answers from the same unified log the agent operates on, so context and execution live in the same brain.

Instead of handing work back to a human after producing an answer, Dot deploys specialized sub-agents to coordinate operational tasks across the broader GTM stack. That makes Default less of a chatbot platform and more of an execution layer for revenue teams.

Learn more about Dot, Default’s RevOps AI agent.

Built for modern RevOps architectures

Default isn't a replacement for Agentforce in a Service Cloud environment.

If your primary goal is AI-powered customer support within Salesforce, Agentforce and Data 360 remain a compelling choice.

But if your biggest challenge is coordinating revenue operations across multiple systems, a Salesforce-centric customer data platform may not address it.

Most modern revenue teams don't operate entirely within Salesforce—they work across a GTM Frankenstack of CRM, marketing automation, data warehouses, enrichment providers, sales engagement tools, and workflow platforms. In that environment, AI agents need shared context that extends beyond a single application.

Default is designed for organizations that want AI agents to operate across the entire revenue tech stack. And our customers find great value in this consolidation:

Choosing the right AI foundation for your team

The next wave of RevOps AI won't be won by whichever team buys the best agent. It'll be won by whichever team's data foundation makes their agents actually useful.

Before investing, consider where your customer and revenue data actually lives, which systems your agents need to interact with, and whether your primary goal is customer service, revenue operations, or both.

Agentforce and Data 360 make that foundation easy for Salesforce-first organizations. For teams whose revenue data lives across the full GTM stack, the same principle applies. The foundation just has to be built for cross-platform reality.

Explore how Default builds the agent-ready data foundation for your GTM workflows. Book a demo today.

FAQs

1. Does Agentforce require Data Cloud to run?

Yes, but not always a paid implementation. Agentforce relies on Data 360 capabilities to provide trusted business context, and most organizations can get started using the free Data 360 allocation included with Salesforce Foundations. A paid tier becomes necessary when you need RAG over unstructured content at scale, identity resolution across non-Salesforce sources, or high-volume segmentation.

2. What's the difference between Agentforce and Data 360?

Agentforce is Salesforce's AI agent platform. Data 360 is the customer data platform that unifies customer information, resolves identities, and provides trusted context for those AI agents.

3. Is Agentforce only useful with Salesforce data?

No. Agentforce can use information beyond standard CRM records. Through Data 360, organizations can connect external data sources, unify customer identities, and retrieve relevant business context from both structured and unstructured enterprise data.

4. What's the biggest consideration before implementing Agentforce?

Data readiness. Most implementation challenges stem from fragmented customer data, duplicate records, and governance rather than the AI itself. Organizations with clean, unified data typically realize value from Agentforce much faster than those trying to fix data quality during deployment.

5. How much do Agentforce Data 360 credits cost?

Flex Credits are $500 per 100,000 credits, with standard actions at 20 credits ($0.10) and voice actions at 30 credits ($0.15). Data 360 has a separate credit pool with operation-specific rates.