Identity resolution

What is identity resolution? 

Identity resolution is the process of finding the unique identity of a customer while creating a unified customer database. It’s done by linking all the customer engagement data and matching it to the right identity of the customer, avoiding duplicates, messy integrations, and wrong targeting. Identity resolution connects customer identifiers like contact details, email, social profiles, feedback and reviews, and other product engagement data – all gathered from different platforms. The main use cases of identity resolution are customer 360, customer journey mapping, personalized & targeted marketing campaigns.  

Some popular identity resolution tools are Segment (a part of their CDP solutions, Zingg (identity graph & entity resolution), Tealium, OneTrust, and more. You can also build custom solutions to suit your data landscape requirements. 

Identity resolution in CDP 

Identity resolution in CDP is a crucial process required to unify customer data. It ensures fragmented pieces about a customer data and form an all-inclusive customer data profile. Benefits of identity resolution for CDP include: 

Personalized marketing: brands can run targeted campaigns, delivering the right message to the right person. 

Clean, reliable & hygienic data: no duplicates, data inconsistencies, or incomplete data that happens while aggregating data from various sources. 

Accurate insights: you could get 360-degree view of customer & their interactions with your brand. 

Cross-channel consistency: customers interact with the brand through multiple channels; & still could receive consistent experiences. 

Similarly, identity resolution in MDM (master data management) ensures data consistency, accuracy, and uniqueness of customer profiles – be it customers, employees, or suppliers. 

The industries that are benefited the most by identity resolution include retail & eCommerce, healthcare, telecommunications, media & entertainment, and travel – all the industries with complex customer journeys and that require personalized messaging and reach out, fraud mitigation, & compliance care. 

How does identity resolution work? 

Identity resolution is what you need to get a single, unified customer record system. Here's how it works. 

Data collection 

Collects data from one or multiple customer interaction platforms. (Social media, CRM, website data). It could also be other offline data like store visits, purchases, etc. Some industry deal with data from third parties, which is also applicable under this approach (data shared by partners, event, loyalty programs, etc).  

Through API and data pipelines, these data are collected in real time or near real time. 

Data cleansing & standardization 

Oftentimes, data from different sources come in different formats. Example: date being written as dd/mm/yyyy & mm/dd/yyyy. Hence, this step involves standardizing and formatting data into unified format and removing duplicates & any inconsistencies. 

Data linking 

This is where things happen – recording every identifier point to each individual. Identity resolution uses two approaches for data matching/linking - deterministic matching and probabilistic matching. Deterministic matching (exact match) is like basic level that looks for unique identifiers like email addresses or phone numbers and matches the ones with similar data. If two records have the same email, then it both are the same customer. 

Probabilistic matching (fuzzy match) uses AI and ML techniques – looking for patterns to connect. Example: one user might have checked out the website on phone and later purchased using the laptop based on a few factors (browsing patterns, device ID, etc). 

For a successful identity resolution with 100% accuracy, combining probabilistic matching and deterministic matching is recommended. 

Merging into a profile 

Once identities are matched, then data is stitched together to form the unified customer profile.  

Identity profile creation 

The identifier has a database that maintains all the linked identifiers of each customer record, creating your golden record – frequently updating any new online or offline identifiers.  

An example of identity resolution would be a customer checking out your product through your website, signing up for offers with phone number, and later purchase the product through a physical store. The identity resolution system connects all these data points, ensuring all teams have the same view about the customer. 

Identity resolution vs entity resolution 

Identity resolution is matching individual identities across multiple sources for customers or users. But entity resolution works on resolving identities of any data—people, product, organizations, etc.). hence, entity resolution could be a broader aspect of identity resolution – focusing beyond people. While the key use cases of identity resolution are CDP, fraud prevention, personalized customer targeting, entity resolution’s use cases include inventory tracking, financial record matching, product catalogue matching, and more. 

How is identity resolution different from traditional data matching? 

Traditional data matching involves heavy manual work, where datasets are checked against pre-defined rules. Even with slight variations in inputs or formats could cause heavy errors. Identity resolution is different from this, has the flexibility to adapt to dynamic environments, de-duplicates and matches in real time, and creates accurate output.  

To sum up, 

With more ways to access a company information or make a purchase, customers are all over the place & so as the digital trail they leave. It has become complex to understand customers, let alone appeal to them with targeted messaging or recommendations. Hence, identity resolution is crucial for businesses to maintain standard customer/user records, so they could grasp their behaviours & expectations, enhance experience they receive, all while staying compliant to data regulations like GDPR.