Most practical uses of Data 360, the product Salesforce previously called Data Cloud, fall into five groups: showing a unified customer profile on the records staff already use, building segments for marketing, computing calculated insights such as lifetime value, grounding Agentforce agents in data beyond the CRM, and triggering flows when unified data changes. For a first project, enriching CRM records or building one segment usually pays off fastest, because both test your identity matching before an agent or automation depends on it.
A note on the name
Salesforce now markets the product as Data 360, and many screens, help articles and contracts still say Data Cloud. They are the same platform. This guide uses Data 360, and the feature names below follow current Salesforce documentation, which may shift again; check the names in your own org before writing them into a scope.
Use case 1: a unified profile where people already work
The least glamorous use case is often the most useful. A service rep opening a contact sees recent orders from your commerce system, open invoices from finance and last week's product usage, without switching screens or asking a colleague to run a query. Data 360 does this through enrichments. Related list enrichments show unified data as a related list on a standard or custom record page, read live from Data 360 rather than stored as a copy in the CRM. Copy field enrichments bring selected values, such as a customer tier or a total, onto fields of the record itself, where they can be used in reports, list views and page logic.
This is a good first project because the audience is internal and the feedback is immediate. If identity resolution has joined two different customers, a rep will notice within a day. Better that than finding out when an agent quotes the wrong order history to a customer.
Use case 2: segments built on behavior and transactions
Segments are audiences defined over unified data: customers who bought a product line but not its accessories, accounts whose usage fell last quarter, or subscribers who opened messages but never registered. Because the segment draws on every connected source, marketing can target on behavior that the CRM alone never held. Segments are then activated, meaning published to the Salesforce apps and channels that will use them, such as Marketing Cloud, and refreshed so the audience stays current.
The risk with segmentation is building dozens of audiences nobody sends to. Start with one segment tied to a campaign that already has an owner and a budget, and measure it against the list the team would have pulled the old way.
Use case 3: calculated insights
Calculated insights are metrics computed across unified data, defined once and reused everywhere. Typical examples are customer lifetime value, days since last purchase, support cases in the past ninety days, or a product adoption score built from usage events. Salesforce also supports streaming insights for signals that need to update in near real time. Once an insight exists, it can drive segments, appear on CRM records through enrichments, and serve as the condition that starts a flow.
The hard part is rarely the calculation. It is agreeing on the definition: which orders count toward lifetime value, whether refunds are netted out, and which system is authoritative when two disagree. Settle that in writing before anyone builds the insight, or each team will keep its own version in a spreadsheet.
Use cases 4 and 5: grounding agents and triggering flows
Agentforce agents answer well only when they can see the right context. Data 360 supplies it in two ways. Structured data, such as unified profiles and insights, gives an agent facts about the specific customer it is helping. Unstructured content, such as PDFs, call transcripts and internal wikis, can be indexed so that retrievers pull relevant passages into a prompt, the pattern known as retrieval-augmented generation. An agent handling warranty questions, for example, might combine the customer's registered products with the relevant section of a product manual.
Data 360-triggered flows close the loop from insight to action. According to Salesforce, these flows can start when data in a data model object changes or when the results of a calculated insight change. A drop in a key account's usage score could create a task for its account manager; a high-value customer opening a third case in a week could alert a service lead. Design these carefully: decide who owns each threshold, and make sure a flow that fires repeatedly for the same customer does not bury someone in duplicate tasks.
Comparing the five use cases
| Use case | Who benefits first | What it depends on | Relative effort |
|---|---|---|---|
| Unified profile on CRM records | Sales and service reps | Reliable identity matching and a clear choice of fields to show | Low to moderate |
| Segmentation and activation | Marketing | Consent data, a destination, and a campaign ready to use the segment | Moderate |
| Calculated insights | Leadership, marketing, account teams | Agreed metric definitions and complete source data | Moderate |
| Grounding Agentforce | Customers and staff using agents | Governed content, tested retrieval, and a defined agent job | Higher |
| Triggered flows | Whoever acts on the alert or task | Insights or data changes worth acting on, and a named owner for each rule | Moderate, rising with volume |
Which to start with
The five use cases build on each other, so order matters more than ambition. For most organizations, the order below keeps each step resting on one that has already been proven:
- Connect two or three sources that answer one known question, and tune identity resolution against records your team can check by hand.
- Surface the result on CRM records through enrichments, so the people who know customers best can spot bad matches early.
- Add one calculated insight whose definition the business has signed off, and show it on the same records.
- Build a segment or a triggered flow that uses that insight, with a named owner and a measure of success.
- Ground an agent only after the profile, insight and content it relies on have held up in daily use.
Companies that jump straight to the agent often discover their matching rules and definitions in production, in front of customers. Working through the list costs a few extra weeks and avoids that.
What each use case needs before you build it
General readiness, meaning clean CRM data, source owners and privacy review, applies to all five. Each use case also has a prerequisite of its own. Enrichments need a shared identifier present on the CRM record, or the lookup has nothing to join on. Segments need consent and preference data that is current in the source, not reconciled once a year. Calculated insights need the history behind them, so a lifetime-value metric fed only this year's orders will mislead. Agents need content that someone owns and retires when it goes stale. Triggered flows need a person who will act on every alert they produce, or they become noise within a month.
