Customer Data Platform — Deep Dive
Great customer experience comes from aligning data to the customer journey. This starts with defining and developing a unified customer data strategy that includes the customer engagement lifecycle stages— Acquire, Engage, Monetize and Retain
According to published definitions by the Customer Data Platform Institute:
A CDP is packaged software that creates a persistent, unified customer database that is accessible to other systems. This definition has three critical elements:
- Packaged software: the CDP is a prebuilt system that is configured to meet the needs of each client. Some technical resources are required, but it does not require the level of technical skill of a typical data warehouse project. This reduces time, cost, and risk and gives business users more control over the system
- Creates a persistent, unified customer database: the CDP creates a comprehensive view of each customer by capturing data from multiple systems, linking information related to the same customer, and storing the information to track behavior over time. The CDP contains personal identifiers used to target marketing messages and track individual-level marketing results
- Accessible to other systems: data stored in the CDP can be used by other systems for analysis and to manage customer interactions
Customer Data — many types, forms and uses
CDP’s are powerful at harnessing 1st & 3rd party data on customers to for a holistic view; but it won’t solve for everything
CDP Core Capabilities
- Omni-channel data collection: Collect as all 1st party data (all customer interaction data). Mobile, web, digital advertising, email, social and IoT devices. Plus, offline data like CRM, POS and Loyalty.
- Identity resolution, Unified profile: Create and manage actionable customer profiles. Unified profiles through “stitching” across devices and channels to ensure one single Guest identity
- Privacy and Security: Identify privacy risk, control data leakage and shut off unauthorized website trackers. Data governance capabilities and flexible opt-out policies for Guest control and evolving privacy regulations
- Tag/Event Management: Enable turnkey integration with hundreds of 3rd party technologies. This allows for creation of a Guest data layer that unifies and standardizes data across multiple applications
- Web/Mobile Personalization: Target products, offers, messaging based on near real time customer interactions across channels
- Marketing Activation: Create and activate segments for different marketing channels for personalization and hyper-targeting
- Customer 360 Analytics: Deep behavioral analytics with transaction and event data. CDPs are not a BI platform per se, but fuel insights with a robust, clean Customer 360
CDP empowers the customer data journey to drive deeper understanding
Using a data driven approach unlocks greater value and opportunity for both the brand and the customer
CDP sample use cases
There are multiple use cases across different business opportunities that get enabled through CDP capabilities
- Increase new sales — Segment cart abandoners, lead management, prospecting look-a-likes with paid media
- Increase engagement / Decrease cost to serve — Post-purchase/ registration targeting, universal suppression across channels, customer service follow-up, measure CLV
- Drive cross/up sell — Target based on in store/branch/online behaviors, personalized offers, product recommendations
- Improve retention — Loyalty segmentation, contract renewal promotion, churn identification & segmentation, re-engage inactive customers
Identity Matching Spectrum — Deterministic to Probabilistic
There is no single view of the customer. Identity is a spectrum. For different use cases, there could be varying certainty — for example to view my orders, specific pii information would be needed that would make the identity deterministic however for personalized recommendations there would be multiple sources stitched together giving a high probability for identity match.
The identity match process broadly follows these steps:
- Hygiene — Source data is decrypted and standardized, bad data is suppressed (test data, toll free numbers, etc), standardized data is encrypted
- Candidate Selection — Existing and historical records possibly related to incoming data are identified as candidates
- Match — Incoming data and candidates combined into single stream, Individual, household, and address matching
- Load Universe — Current, historical, and cross-reference tables loaded
Sample Architecture
Specific use case to develop a unified data platform that serves 3 customer databases: marketing, deterministic and analytical
Data Ingestion/Processing
The data architecture follows the bronze, silver, gold data engineering paradigm with the following separation of responsibilities:
- Bronze layer — Data acquisition raw data to data lake
- Silver layer — Clean and de-dupe, hygiene rules, match logic
- Gold layer — Single customer view or customer aggregate/golden record, various attribute subsets including analytics, derived
Data Model
There are multiple data models and associated schemas applicable based on the data domains or subject matter area. For example, profile and preferences would be part of the Deterministic Guest database. Single customer view or customer golden record would be part of the Marketing database
CDP Implementation Steps
A generic CDP implementation follows a sequence of steps. These would vary based on use cases, prioritization, timeline, resourcing and cost
In Closing
Meeting the demands of consumers requires harnessing the power of data & insight across an expanding volume of complex consumer data. CDP enables brands to place the customer at the center of the experience thus unlocking new value and promoting a culture of data, agility and innovation
