Identity resolution is identifying, cleaning, matching, merging, and connecting fragmented information about an individual across various platforms, databases, and devices to form a comprehensive, accurate, and up-to-date profile. This process integrates data from sources like marketing technology, enterprise systems, and data lakes to create a complete view of a customer. In identity resolution, “customer” refers to a household, prospect, patient, Employee, or even a product, depending on the business purpose.
Importance of Identity Resolution
Identity resolution enables businesses to distinguish one customer from another, even with potentially conflicting data points, such as email addresses, device IDs, or physical addresses. By connecting various identifiers like emails, phone numbers, and transactions to a unique customer ID, businesses can build trustworthy profiles essential for personalizing customer experiences. This process is central to creating a consolidated customer view, sometimes called a Golden Record, which stores comprehensive information about an individual, enhancing customer insights and engagement.
Golden Record Explained
A Golden Record is a unified view of a customer, integrating data from all possible sources, such as websites, CRM systems, and social media. This record consolidates identity graphs, transactional data, preferences, and contact history. Advanced identity resolution and data quality measures, like real-time updates and precise merging, power these Golden Records, forming the basis for delivering personalized customer experiences.
Deterministic vs. Probabilistic Matching
Deterministic and probabilistic matching are two methods used in identity resolution:
- Deterministic Matching: This method links records with identical identifiers (like phone numbers or email addresses) across different systems, ensuring an exact match. For example, it matches two profiles with the same email address across devices.
- Probabilistic Matching: Uses analytics to identify records that likely represent the same individual despite minor identifier differences. For example, detecting that “David” and “Dave” may refer to the same person based on contextual similarities. This method provides flexibility, addressing human error or alternate entries and allowing the formation of more complete identity profiles.
Levels of Identity Resolution
Identity resolution has different levels based on the purpose:
- AdTech Identity Resolution: Primarily uses third-party data, often with encrypted identifiers, for general audience targeting.
- MarTech Identity Resolution: Focuses on first-party data, using advanced methods to create precise Golden Records for consistent, personalized customer experiences.
- Regulated Industries (Healthcare and Financial Services) require highly accurate identity resolution, often involving exact matches, to safeguard sensitive information.
Role in Customer Data Platforms (CDPs)
A customer data platform (CDP) should ideally support identity resolution, enabling the creation of unified profiles. However, not all CDPs offer advanced capabilities. Basic CDPs may rely only on deterministic matching. In contrast, advanced CDPs support deterministic and probabilistic methods, providing flexibility for managing multiple identifiers and potential data inconsistencies.
Householding in Identity Resolution
Householding refers to grouping related identities, such as family members in a shared household or members within a B2B organization. Advanced CDPs use householding to form relationships between individuals (e.g., household members) or between businesses, maintaining separate identities while linking related records. This process provides deeper insights into group behaviours and interactions.
Critical Components of Identity Resolution
- Data Quality: High-quality, validated data is foundational for effective identity resolution, ensuring accurate matching and trustworthy results.
- Data Governance: Ensures data is used according to organizational policies, addressing privacy, security, and permissions.
- Data Ingestion: The process of importing data from various sources, mapping it to existing attributes, and preparing it for analysis. It includes structured and unstructured data from CRM, IoT, and social media sources.
- Data Matching and Merging involves linking related records across data sources and merging them according to business rules to form a single identity while managing distinct identifiers.
- Persistent Key Management: This system manages unique keys for identifying customers across multiple records and interactions, enabling a longitudinal view over time.
Identity resolution, especially within a CDP, is essential for creating reliable customer profiles that inform personalized experiences and effective business decisions. By accurately linking data from multiple sources, businesses gain insights that drive customer-centric strategies and improve overall engagement.
