One of the most important objectives of Data Cloud is data harmonization, which is achieved through data standardization and unification.
In this context, the concept of a Unified Profile and its associated terminology, such as the golden record and key rings, becomes extremely important.
Identity Resolution and the Unified Profile
The process of creating a unified profile begins with identity resolution and addressing the variations a single customer might exhibit across different platforms. For instance, a customer named Michael might also be known as “Mike” or “Micheal” and could have multiple addresses due to relocations. Traditional deduplication might miss these nuances, but resolving such discrepancies is a core strength of Data Cloud.
It’s crucial to understand that Data Cloud doesn’t merge records; it compiles them to create a “golden record”. The original records remain intact in their source systems. The golden record serves as a comprehensive view of the customer, combining data from various sources.
This is achieved through match and reconciliation rules, utilizing both deterministic and probabilistic matching to account for different ways individuals represent themselves in data.
The Golden Record
The golden record provides a 360-degree view of the customer, often referred to as the “source of truth”. Survivorship rules determine which values are retained in the golden record, ensuring consistency for all who access it.
Essentially, this method selects the best option to unify data into a single, simplified view of the customer.
The Key Ring Analogy
Contrasting the golden record is the “keyring” method, also known as the “key chain” or “bucket” method. This analogy helps illustrate that. Like a key ring holding multiple keys, a Unified Profile in Data Cloud retains various pieces of information about the same individual.
Rather than selecting a single “best” option, it summarizes all relevant data, avoiding data loss since no records are merged, deleted, or overwritten. The right data is used for the appropriate context, like using different keys for different locks.
The key ring method aligns with the decentralized data reality many teams face, preserving data access without disrupting existing structures. Maintenance is less complex, relying on identifier mapping rather than survivorship and consolidation logic. This approach offers better adaptability to changing data models, simplifying management and updates.
In essence, the key ring system acts as a System of Reference by linking Source System IDs to create a Unified Profile, rather than consolidating data into a golden record.
Addressing Data Duplication
At this point, you might be wondering with the source data saved in various forms, how to deal with the data deduplication and data storage overall. A recent advancement in handling data duplication is the zero-copy network, introduced this year.
This system allows companies to share data across platforms by referencing it rather than copying it, creating virtual views of data without storing it on Salesforce. This innovation, supported by various partners, can reduce costs associated with data storage, deduplication, and unification processes.
Final Thoughts
Data Cloud offers a robust method for creating a unified profile, enabling a comprehensive 360-degree view of the customer (which isn’t big news since that’s one of the USPs). The key ring method provides a more holistic and flexible approach to understanding and identifying customers to facilitate personalized relationships.
However, practical considerations such as understanding data sources, managing consents, and complying with GDPR or CPPA regulations must be addressed before fully embracing data unification. A clear data model and the right team are predecessors of the successful implementation of data unification solutions.