How to unlock Business Value with Data Cloud and AI

Last week, John Banian and myself had the incredible opportunity to present not one, but two sessions at the Salesforce World Tour. This is a yearly event that we organize at several locations around the world and brings together trailblazers to share insights, discuss trends, and explore the future of different Salesforce products.

In our session, we had the chance to dive deeper into the subject how to unlock business value with Data Cloud and AI, sharing insights from our experiences with Data Cloud implementations that we believe will drive the successful use of Data Cloud forward.

In this article, I will try to share some key takeaways from our session. Whether you’re a experienced Salesforce user, a prospective customer, or simply interested in in Data Cloud, I hope you will find these reflections interesting and useful for your Data Cloud implementation.

Move to Data-driven companies

Today, we know that data-driven companies are outperforming their competition. But what does it mean to be data-driven? At Salesforce, we define this as anyone at your company having access to the right data and the ability to use that data to unlock insights and make informed decisions.

However, data management is complex – and having access to customer data has not been getting any easier. In order to achieve success now, companies need to learn how to harness this data and use it to build exceptional customer experiences. Doing this takes a tremendous amount of time from your IT teams. 36% of IT's time is spent designing, building, and testing custom integrations across multiple systems (Source: MuleSoft Benchmark Report, 2022). These integrations are expensive to maintain, they're brittle, and maybe most difficult of all they must follow the growing security, privacy, and compliance regulations emerging in the world.

And data is important because it’s the fuel that powers the AI engine... but also exposes some risks if not leveraged in the right way.

At Salesforce Professional Services, we have successfully completed over 300 Data Cloud implementations globally across various industries. From these real-world experiences, we have distilled the following five pillars around best practices and principles for implementing Data Cloud that will lay the foundation for building your Data and AI capabilities.

Business Strategy, Data Strategy & Governance

As you start with the end goal in your mind - We always first define the Business Strategy

1. Formulate the business vision for Data Cloud - prioritise use cases and first focus on foundational capabilities.
2. Define measurable Success Metrics - priorities may change but they should be easy to measure.
3. Gain commitment from Business, Data & Tech Teams - to solve the last mile problem of data, they have to be on the same page.
4. DC is a business user platform for combining Data & Democratisation. Plan use cases that lay the foundation for AI models.

Once you have a good idea about your Business Strategy, you can start to focus on Data Strategy & Governance.

1. Understand Enterprise Data Lifecycle, Lineage and compliance needs to work on data model and integration.
2. Marry your Data framework & Tech capabilities - Align your MDM with the Business Warehouse as well as your Data Lake. Know Data Cloud’s capabilities and how it can work together with other systems in your landscape.
3. Map all data sources and their relationships with a lens on the business use case. Bring only fit-for-purpose data into Data Cloud. You can always add more data at a later stage.

Technical Optimisation

Data Cloud has a Pay per use model and the platform offers multiple ways to design your required solution. Set up Data Cloud with an experienced team to run it optimally.

• Review performance and scalability of each design option with a lens of ROI - same thing can be done in many ways - Familiarise yourself with the Consumption model before finalising the architecture.
• There are multiple ways to Integrate, calculate and activate data. And if you want to abstract yourself from the hassle of traditional ETL, leverage Zero-Copy integration and optimise on multiple fronts - why store data locally when you can efficiently access it on the fly?
• And lastly, stay up to date with the product releases to make sure you are always using the latest and greatest coming from our Product Team. This can benefit with your own enterprise roadmap and also helps avoiding potential tech debt.

Adoption & Trust

Any IT solution is only as good as how it is adopted by Business teams.

1. To ensure good adoption perform Business Process Change review and plan for teams to level up their knowledge to establish comfort.
2. Run demos and guided user sessions. Draft Manuals for common use cases and share ways to learn offline + How to ask for help. Data Cloud is a very intuitive tool and it won't be long before Business users start teaching tricks to IT.
3. Remember the Success metrics we defined at the Business Strategy step? Report on these KPIs to track the success of your implementation ensuring wider business adoption.
4. Throughout the process have Business & Tech SMEs embedded in projects to ensure project bearings are auto-corrected.

All this setup is to manage end customer’s data and Trust is vital

1. Trust is one of our core values. Data Cloud is built on the industry's most trusted infrastructure.
2. Ethical data usage for AI is all about the transparency how data is used and how decisions are made. Besides that it requires clear accountability for the outcomes of AI decisions.
3. Salesforce Data Cloud is compliant with industry regulations such as data privacy, GDPR and cyber security. You can trust Salesforce with your data.

Once you have the data idea in order, you can continue with AI. Remember the five pillars of success we just discussed? With these in mind, we want to help set you up for success so you can create AI-powered experiences that you and your customers will love.

1. First, you need to build your source of truth. This can be difficult, but you start small and build out over time. Think of the data you need, and especially what is the quality of this. Connect the channels you want to use, whether it is a chat bot or a knowledge center.
2. From there you can launch your first AI journey to set the foundation. This means creating AI models and testing new use cases for your AI. Start small, and add more use cases over time.
3. Now you’re ready to create AI-powered experiences, like generative replies, recommendations, and search! And your service teams are key to the formula here: There is always a human in the loop, giving feedback, reviewing and approving the output of Einstein copilot before it gets deployed.

This is how you’re going to create scalable experiences that every customer can trust.

Customer successes and challenges

During the sessions we also shared some customer successes and challenges. Looking at the different successes and challenges we experienced during implementation projects, we can see that it was either because customers stuck to the above pillars or actually lacked some of them. Customers that had success with Data Cloud had a clear vision and aligned success criteria and use cases, accepted that some of their requirements were not a proper fit to Data Cloud (yet) and made sure there was proper alignment between different teams internally. The customers that had challenges, had big challenges with their use cases, misaligned management teams or had bought Data Cloud without first checking whether their requirements were a good fit.

Try and stay close to the pillars I just discussed and you can also have success with Data Cloud.