What is Salesforce Data Cloud?

71% of company applications are disconnected – Parker Harris, Co-Founder Salesforce

Customer360

The promise of a customer data cloud—Customer360—has always been the north star for Salesforce. For most organizations, customer data is siloed in multiple systems, not just Salesforce, and it is often codified in different formats. Salesforce Data Cloud provides a single view of customer relationships with clicks, not code, harmonizing data (unify, manage, cleanse) from disparate systems. That data is visible so that low-code and pro-code tools like Flow and Apex can provide actionable insights.

The Salesforce Platform, Data Cloud, and Einstein are fully integrated and are now called Einstein 1. What makes this integration possible is the metadata framework that underpins Salesforce. It connects Data Cloud with the core platform and AI. Metadata is a huge differentiator. It drives configuration and makes the low-code experience possible. It makes the Data Cloud integration simpler. But it also enables AI to deliver better results.

“Einstein Copilot understands your business. You’ve taught Salesforce your business with metadata –  with custom fields. It knows that there is a “High Value” custom field. It interpreted that.” – John Kucera, SVP Product Management, Salesforce.

Connecting data

What makes Data Cloud powerful is it enables you to connect your different data sources, harmonize the data and then make it available to the core platform. That data can be copied into Data Cloud, or it can be Zero Copy for data lakes like Snowflake, Google, Microsoft or AWS.  Zero Copy means that the data is not copied and stored in Data Cloud, but it is available, making it a real-time data platform. For many of your data sources, a subset of their data is copied in near real-time into Data Cloud through easily implemented integrations. The ability to connect any data source using clicks, not code, dramatically reduces the development effort involved in getting an aggregated view of your data. Then, you can drive actions and enhanced workflows from that data.

“Data Cloud is the heartbeat of the Salesforce Platform.” – Rahul Auradkar, EVP & GM of Unified Data Services & Einstein, Salesforce.

Extending the metadata framework

We will get into the details of this later, but at a high level, the data is stored in objects called Data Lake Objects (DLOs) that have fields. For Zero-copy data sources, the DLO is a reference to the data location. Data Model Objects (DMOs) are virtual objects that are related to DLOs and are used to harmonize and normalize the data.

There are over 100 standard DMOs that you can extend with custom fields, and you can also create custom DMOs. When you are building Flow or writing Apex, you can reference the DMO fields, which are visible as metadata on the core platform. The magic of the metadata platform is that you treat DMO fields like any other standard or custom metadata.

For organizations with data-driven teams, the clicks-not-code approach will dramatically reduce development and maintenance effort. However, for teams more focused on core process-focused Salesforce functionality, there is a learning curve. Whilst this is not huge, it should be factored into the initial implementation timescales. 

History of Salesforce Data Cloud

Salesforce Data Cloud has undergone multiple name iterations: from Customer 360 Audiences to Salesforce CDP, then to Marketing Cloud Customer Data Platform, followed by Salesforce Genie, and most recently, Salesforce Data Cloud.

Some of these name changes were due to names that didn’t resonate, but more importantly, many of the renamings reflect a significant evolution in the product. 

• Customer 360 Audiences: Salesforce’s first CDP offering was introduced in 2020.
• Salesforce CDP: Renamed in 2021 to align with industry terminology as the CDP market began to flourish.
• Marketing Cloud Customer Data Platform: In 2022, this new name emerged as part of Salesforce’s effort to simplify its marketing product names.
• Salesforce Genie: Also in 2022, at Dreamforce, Salesforce introduced Genie, marking a shift in use cases beyond marketing to include sales, service, and more, along with a new zero-copy architecture.
• Salesforce Data Cloud: In 2023, Salesforce dropped the name Genie but kept the mascot, rebranding it as Data Cloud and aligning it with Salesforce’s GenAI innovation.

The development of this CDP offering spanned several years, with significant development investment before it was brought to market. These name changes were not merely cosmetic but indicative of substantial product evolution. Salesforce Genie/Data Cloud represents a significant leap forward, transforming the CDP from a marketer’s tool focused on deduplication and customer profile matching into a versatile platform that supports sales, service, and other functions across the Salesforce ecosystem. Salesforce has built Data Cloud to support multiple use cases, making it a valuable resource for any team. However, its marketing genealogy is very clear, and currently, the strongest use case is providing a unified view of the customer.

How Data Cloud has revolutionized the digital landscape

Data Cloud enables Salesforce to ensure that customer-related data from these other apps are able to enhance the data held in Salesforce to give a more insightful view of the customer. This makes the data more actionable for AI and automation. Recent research by Salesforce has revealed that customers have over 1,000 enterprise applications used by different departments – marketing, sales, support, product management, delivery, finance, etc. Many of these hold customer data in different formats and use different approaches to identify the unique customer. For example, email, cellphone, a generated customer number, a customer-created username, or SSN.

This makes it very difficult to build a unified view of the customer that all systems and departments can agree on and access. This siloed approach makes the customer experience frustrating and disjointed from the customer’s perspective. Customer satisfaction suffers. It makes it inefficient and error-prone for the organizations’ department to communicate. Productivity drops. It also makes it impossible to generate accurate AI-driven insights from Einstein AI. Failing to leverage AI is a major misstep.

Every organization is striving for a single view of the customer so that they can market, sell, and service them in a way that is highly targeted and personalized. 

Before Data Cloud existed, the only approach was to build a number of highly customized native integrations between every system to extract and harmonize data. And every time a core system was changed, the integrations needed to be updated. Whilst not impossible, it was cost-prohibitive.

Data Cloud makes it possible by providing an approach that is transformative. It is a metadata-driven centralized data hub. It comes with pre-built connectors to a huge range of systems eliminating the need for custom code. It has the tools to ingest, harmonize, and unify data. The data is made visible to Salesforce via custom metadata. This is what sets Data Cloud apart from other “data lake” applications. This allows the data to be exposed in the Salesforce UI, analyzed via AI, or used in automation. This is all configured through a familiar drag-and-drop Salesforce metadata framework, which is understood by the huge ecosystem of Salesforce professionals.

If organizations already have sophisticated Data Lake implementations, Data Cloud is able to treat them as external data sources without copying data. But Data Cloud can also ingest and store subsets of data where necessary. By combining both approaches, it ensures it does not make the duplication of data worse. 

As every organization strives to provide a more personalized customer experience, Data Cloud provides an approach that is affordable and maintainable, no matter what combination of enterprise applications has been implemented. It ensures that Salesforce remains the core customer-centric enterprise application in the organization’s digital landscape. 

How does the Data Cloud work?

Data Cloud is part of the Einstein 1 Platform, which helps you and your team gain a complete understanding of your customer based on data. The platform is constantly updated with releases every 2 weeks, rather than the “3 Releases A Year” that we see with the core platform. Therefore it is rapidly evolving and becoming more enterprise-ready every release.

There are several steps to get from multiple data sources to data that can be visualized and actionable inside Salesforce. Each step has detailed functionality in Data Cloud. The Data Cloud Consultant Certification will ensure you understand how to use the functionality to build Data Cloud. Spoiler alert: It is 80% planning and 20% building. So the Cert will help you with the 20%.  We have a proven approach, content and resources that will help you with the missing 80%.

Here is a summary of the Data Cloud step and the functionality provided to support each step. As you can see, this is a complex product because it is doing something complex. What Data Cloud has done is hide the configuration complexity. But you still need to know what Data Cloud is trying to achieve so that you can use the right functionality based on the shape of the data and the desired end result.

Connect: A unified platform for seamless data integration.

Effortlessly integrate all your data with Salesforce, no matter the source. With Data Cloud, you can unify disparate data without the need for costly and complex data pipelines or discrete integrations.

Support for any source across your enterprise: MuleSoft offers built-in connectors for easily importing data from leading cloud providers, web/mobile applications, and over 300 sources. 

Modern Data Lakes and Warehouses: Securely share data between Data Cloud and leading data lakes or warehouses with zero-copy integrations, eliminating data duplication and reducing storage and ingestion costs.

Ingest unstructured data: Expand beyond structured data to include unstructured content like chat transcripts, PDFs, and knowledge base articles with Data Cloud’s Vector Database.

Salesforce Clouds and Orgs: Consolidate data from all your Salesforce orgs and clouds for a complete customer view across your Salesforce ecosystem. Utilize near-real-time sync for CRM-related applications to act on data instantly.

Harmonize: Turn disparate data into a unified model.

Data Cloud harmonizes all your data into a comprehensive model with a 360-degree customer view. This gets all the data in a state that it can then be manipulated in Salesforce objects.

Reference Data Models: Transform data from any source into Salesforce-native objects and fields using the standard metadata framework. Start quickly with out-of-the-box reference data models or fully customize them to suit your needs.

Build and deploy Data Graphs: Make specific portions of your enterprise data model available for services like AI to improve speed and performance. Use Data Cloud’s visual builder to view and manage relationships between data model objects before deployment.

Unify: Achieve a complete view of each customer.

Unlock a 360-degree view of every customer that’s accessible and understandable by all your customer-facing teams. This is a critical step as customer data often has different identifiers, leading to duplication.

Matching Rules: Unify profiles from various data sources by applying rules during the identity resolution process. Choose between Exact or Fuzzy Matching rules to meet your precision requirements.

Reconciliation Rules: Select which field values to use on unified profiles when conflicts arise from different data sources. Choose criteria like Last Updated or Most Frequent to guide your decision.

Manage Multiple Rulesets: Combine matching and reconciliation rules into rulesets to resolve identities into unified profiles tailored to your business needs.

Enhance: Deepen your understanding of customers.

Transform all your data into actionable insights with low-code tools that help you analyze and predict customer behavior.

Calculated Insights: Create valuable metrics like Customer Lifetime Value, Propensity to Buy, and Engagement Scores using our low-code builder.

Segmentation: Identify, analyze, and discover high-value audience segments in an intuitive interface.

No-Code ML Models: Easily connect external models, like those from Amazon SageMaker, to your data in Data Cloud.

Activate: Turn unified data into action.

With a familiar drag and drop interface you can use data in the Data Cloud objects inside Salesforce using tools that your teams have already mastered: UI, Flow, Apex or AI.

Org & Object Enrichment: Enhance Salesforce apps by integrating Data Cloud fields, objects, and insights directly into your contact, lead, and account pages.

Automation & Workflows: Trigger automation using Flow based on any changes within your unified customer data model.

Omni-channel Activation: Utilize your unified data across a growing ecosystem of third-party platforms, including advertising partners like Google, Facebook, and Amazon, as well as messaging platforms like WhatsApp.

Einstein AI: Use the data to drive more insightful sales or service recommendations using it in EinsteinAI prompts in the context of Salesforce customer data. 

AI: Securely ground AI in your data.

Use the Einstein 1 Trust Layer to securely leverage all your structured and unstructured data for AI, maintaining data governance and security.

Ground AI Prompts: Provide AI prompts with the proper context from your structured and unstructured data and its metadata within Data Cloud.

Einstein Copilot: Generate accurate and trustworthy natural language responses grounded in your company’s proprietary data stored in Data Cloud.

Final word

Data Cloud aggregates customer data that lives in external siloed applications. It provides the 360-degree view of the customer that has been a promise for the last 30 years. ​​It enables insights and actions to be driven from within the Salesforce core platform using clicks, not code. But, this added complexity means that success requires a more rigorous, intentional, and architected approach to implementation. 80% of the implementation project is planning, and only 20% is configuration.