Salesforce Data 360 implementation and activation

Turn a unified profile into an accountable action

Customer information can exist in several systems without being useful at the point of decision. A service adviser cannot see the relevant purchase, marketing selects an outdated audience, or an assistant receives incomplete account context. Emerge designs Data 360 around the action that needs better information and the source records required to support it.

Data 360 is Salesforce’s current name for the product formerly called Data Cloud. Existing contracts and enabled capabilities must be reviewed in their own terms; a naming change is not a reason to assume every new feature or connector is included. We confirm the customer’s packaging, consumption arrangements and environment before specifying implementation scope.

Define the customer model before connecting everything

The first task is a source and purpose inventory. We identify which system owns contact details, account relationships, transactions, service history and consent. Each proposed field has a destination use, expected freshness and a business owner. That limits unnecessary ingestion and exposes disagreements that a technical connector would otherwise conceal.

Identity resolution requires explicit rules and review. Two records with the same email address may represent one person, a shared mailbox or a reused address. A customer can also use different identifiers across retail, service and business purchasing. We test matching and reconciliation on representative examples, including ambiguous cases, rather than treating a larger unified audience as evidence of better quality.

Semantic definitions matter as much as matching. A purchase, order, refund and active customer must mean the same thing to the teams using a segment or insight. We document the analytical grain and the treatment of historical changes so a later correction does not silently alter the interpretation of a campaign or service decision.

An illustrative data-to-activation workflow

An approved source publishes customer and transaction data through the selected connector or interface. The implementation validates identifiers, schema and required fields before mapping records into the agreed model. Identity rules create a usable relationship between source records while preserving the ability to trace each contribution.

A segment expresses a concrete business requirement, such as customers eligible for a reviewed follow-up. Eligibility includes the required consent and exclusion rules. The activation publishes only the fields and identifiers needed by the receiving channel, CRM workflow or service application. A destination acknowledgement and reconciliation check confirm that the intended audience actually arrived.

An AI use case can consume approved context from the same foundation, but receives its own access and action boundary. A unified profile is not permission to disclose every source attribute. The assistant’s user, task and destination determine which information may be retrieved and whether a proposed action needs review.

Choose ingestion and federation with the actual workload

Salesforce documents both ingestion and zero-copy connection patterns in its data portfolio. Their suitability depends on supported sources, configured capabilities, query behaviour and the use case. We assess the actual combination rather than assuming that zero-copy removes every transfer, latency or governance consideration.

Freshness is specified at the decision point. An hourly audience refresh may be appropriate for a campaign but unsuitable for a time-sensitive service interaction. We trace the delay across source publication, processing, segmentation and destination activation, then test the complete path with known records.

Consumption also follows the configured workload. Data processing, storage and optional capabilities can have different commercial treatment. We establish a baseline using representative activity and map it to business ownership, with the customer’s actual agreement as the reference. No generic credit estimate substitutes for a scoped consumption model.

Validate identity, permission and destination behaviour

Acceptance includes source-to-model reconciliation, reviewed match examples and a sample of records that should remain separate. Tests cover missing consent, a withdrawn preference, changed account relationships and excluded source fields. The receiving team checks that a segment or insight is both technically accessible and operationally useful.

We also test corrections and deletion handling across the selected sources and destinations. A change to the authoritative record should have a documented effect on the unified view and future activation. Historical campaign evidence may have a different retention purpose from the current customer profile, so those responsibilities are clarified rather than merged casually.

Hand over a living data product

Emerge delivers the source inventory, mapping decisions, identity rules, segment definitions and activation contracts. Named stewards own semantic changes and ambiguous matches; platform operators own processing failures and consumption monitoring. The review cadence examines data quality, freshness and whether the activated information improved the intended workflow.

The starting engagement can be one audience or service-context problem with clearly identified sources. A bounded Data 360 release gives the organisation evidence about identity quality, useful activation and operating cost before expanding its customer-data programme.