We understand now that Data Cloud is a powerful CDP tool. To better understand the high-level overview of Data Cloud, we need to understand a few concepts that we haven’t yet talked about in the book so far. Along with the concepts we already discussed in chapters 1 and 2, we will discuss a few more concepts, specific to Data Cloud.
- Harmonization of data
Data harmonization refers to combining data from different sources and providing users with a comparable view of data from different sources. Data harmonization improves the quality and utility of the data. In this context, it would mean bringing the customer data from different sources along with unification, enrichment, and validation in one place.
The net result of data harmonization is to bring together data from different sources into one place so that they function as if they are from a single source.
- Customer graph
Also called an identity graph, refers to a database that stitches customer records from different data sources to create a single unified customer profile. After the process of identity resolution, we create a customer graph.
A customer graph is a framework by which a company determines what type of customers they have, how they are interacting, how the company should interact with the customers, and so on. It enables companies to understand their customers better. Siloed customer channels get unified into one after creating the customer graph (Figure 3.11).
Figure 3.11: Customer graphs help to understand how customers are interacting across channels
In Salesforce Data Cloud, customer graphs are created based on data models derived from their 20 plus year of business experience. Data from different sources are harmonized in a single customer profile.
- Lakehouse architecture
Salesforce Data Cloud stores data using a lakehouse architecture. Lakehouse architecture refers to data lakehouse, which is a new open data management architecture that combines the flexibility, cost-efficiency, and scale of data lakes with the data management and ACID transactions of data warehouses. Thus, it is a combination of the best features of a data lake and a data warehouse. The best of both worlds.
Merging the two into a single system means that customer data teams can move faster as they can use data without needing to access multiple systems. Data Lakehouse is fast becoming an industry best practice for storing customer data (Figure 3.12).
Figure 3.12: Data Lakehouse packs the best of Data Warehouse and Data Lake (source-https://www.databricks.com/glossary/data-lakehouse#:~:text=A%20data%20lakehouse%20is%20a,(ML)%20on%20all%20data.)