Cloud Journey — Part 7 | Customer Data Platform (CDP)

Piece of art by “Gizem Vural”​

Cloud Journey Series:

• Cloud Journey — Part 1. A basic introduction of cloud, applying PACE layering and The 6R’s.
• Cloud Journey — Part 2. A quick review on what is the good organization chart to enable cloud journey.
• Cloud Journey — Part 3. A quick view on Business Values and Business Drivers on a cloud journey.
• Cloud Journey — Part 4. What does cloud mean for your “Talents & Culture”?
• Cloud Journey — Part 5. Using Platform Ops to accelerate DevSecOps adoption
• Cloud Journey — Part 6 | Foundations of Cloud Architecture
• Cloud Journey — Part 7 | Customer Data Platform (CDP)

A customer data platform (CDP) is a technology that unifies your company’s customer data from marketing and other channels to enable customer modeling and to optimize the timing and targeting of messages and offers however there is a misperception that CDPs eliminate friction with customer data hence that is why I am addressing CDP as part of cloud journey. Also there are many occasions that better to and design customer journeys that use customer-directed engagement models to earn (not collect) first-party data for empowering more effective personalization that enhances experience.

I strongly believe the real driver for cloud transformation must be adding value to customers by improving the experience and removing frictions. The prerequisite of such goal is a strong cloud foundation, micro-services, api gateway, APIOps, DevSecOps and customer data platform.

I believe in order to have a successful CDP within your organization first you will need to establish a strong micro-service and cloud foundations (refer to Cloud Journey — Part 6) and an enterprise architecture approach that is best-of-breed driven as such you will be empowered to establish a CDP that is an orchestrator. It is crucial then to integrate your customer identity and access management (CIAM) system with CDP.

Customer data enables continuity across channels, allows to offer personalization, and provides context to analytics and reporting. Supporting growing business demands, especially across an enterprise application landscape littered with customer data silos, is a complex and difficult task. Customer data management is a combination of politics, strategy and technology. As organizations continue to develop their customer data governance strategies, a growing burden is falling on architects to provide a solution that can capture, protect, enable and mine customer data for various requirements. Next, you need categorize the type of customer data initiative in scope. The following figure shows the three categories with some examples of each dimension.

The three categories of customer data management initiatives are:

• Operational: These initiatives are the implementation of technical capabilities that act as shared services across your runtime enterprise application architecture. Examples are customer master data management, global customer ID assignment, consent management, and customer identity and access management (CIAM) services. Customer data implementations in this category require real-time performance capabilities and operational reliability. Requests that fall into this category are either shared services for your overall architecture or modifications to shared data models or integration endpoints.
• Functional: Initiatives that fall into this category are based around the endpoint applications that rely on customer data to enable their functionality. Examples are enterprise applications, such as CRM applications and ERP applications, as well as primarily data-driven applications such as customer data platforms. Components here act as producers and consumers of customer data. Functional endpoint applications will constantly change due to adoption driven by your business requirements across your organization. Enhance functional components by integrating the shared operational services to enable continuity across their customer data silos.
• Analytical: Customer data initiatives that are based on the aggregation of customer data to derive analytics, reports and insights fall into the analytical category. Examples are business intelligence applications, reporting dashboards, and data science or machine-learning-based services that ingest customer data and metrics from your enterprise application architecture. You can then use the output of data-science-related components to provide improved personalization data to your functional endpoint customer application. Use the output of data-science-related components to provide improved personalization data to your functional endpoint customer application. For example, integrate the derived customer segments or customer profile attribution patterns to customer personalization services to improve experience.

Practically speaking, you will have to support multiple integration types to enable the full set of customer data capabilities. Figure below shows an example of the range of integration services required to support each component of your customer data architecture, as well as the ideal integration approach for each component. One of the greatest tools to achieve this is MuleSoft (as your API Gateway) and its powerful tool DataWeave and for events kafka/sns/msk/etc, a simple illustration of your architecture is shown in below:

Event-driven integrations require event broker middleware to enable its pub-sub approach to data flows. API integrations allow for real-time integration calls as well and can flow through an integration middleware, such as an iPaaS, to promote reusability. Bulk ETL data feeds should be leveraged for customer data integrations that do not have a real-time requirement. For example, a nightly update of a data lake or logical data warehouse within your analytical architecture can be done via bulk ETL transfers.

Ideally a CDP should be fed customer data from a diverse set of sources to enhance the customer identity and segmentation outputs. Figure below is an example of the integrated flow through a CDP and highlights the four common functional areas of a CDP.

1. Data ingestion: The ability to ingest first-party, individual-level customer data from multiple sources, online and offline. CDPs vary in data persistence approaches and data engineering enablement.
2. Identity resolution: The ability to consolidate customer identities and customer profiles and connect attributes to defined and derived customer identities, it can be using a simple solution as global customer analytics ID.
3. Segmentation: The ability to create and manage customer segments and segment configuration such as a rule-based segment creation and an admin GUI dashboard for visualization and configuration.
4. Activation: The ability for the CDP to send instructions on how to activate segments for different purposes such as marketing communication, email campaign or commerce personalization services. The downstream applications execute the actual customer communication, dictated by the output from the CDP.

CDPs can perform customer identity resolution based on a deterministic (that is, rule-based) approach, and some CDPs offer advanced probabilistic-style matching (that is, analytics-based) as well. Some CDPs may even allow users to combine the two approaches. Furthermore, some CDPs have advanced functionality to support machine learning (ML) models, either vendor- or customer-provided, for use in custom data configurations. Other areas of advanced or additional functionality that can be found in CDP vendor solutions are:

• Predictive analytics and propensity modeling exposed via dashboards and programmatic endpoints (for example, APIs).
• Next best recommendations made through the use of predictive algorithms — suggesting offers, messages or other actions for the purpose of customer engagement.
• Campaign testing and optimization by integration with tracking and marketing services.
• Customer journey optimization using ingested and serviced analytics and configured customer engagement workflows.
• Customer consent management functionality for tracking and requesting customer data privacy actions (for example, around the right to be forgotten, or opting out or in).
• Domain-specific functionality such as a focus on B2B workflows and business data enrichment, or functionality with an industry focus, such as for the automotive or retail sectors.

CDPs present an array of exciting customer data capabilities. However, a poorly integrated CDP will devolve into just another data silo that provides minimal value to the business.

Following figure shows an example of how CDPs can provide customer journey orchestration options, based on an analysis of customer data. Customer journeys represent the sum of all the interactions and experiences your customers have with your organization, across all your customer-facing channels. CDPs offer customer journey reports driven by ML-enhanced analytics. By using ML algorithms to analyze datasets, these reports offer faster and more comprehensive insights into customer journeys. Marketers and Product Managers can understand which journeys provide better customer experiences and better outcomes, and which journeys are causing customer dissatisfaction. Additionally, recommendations from a CDP customer journey optimization engine can be configured by a marketer and Product Managers to be sent to an integrated activation service, such as a multichannel marketing hub.

The four CDP types (as defined by Gartner) are listed below.

1. CDP Engines and Toolkits: CDP offerings that provide functionality via an open-source toolkit, SDK or set of integratable services. Toolkits are used to build new CDP-focused applications on top of existing data management platforms that contain customer data. Example vendors: aqfer, Jahia (jCustomer).
2. Marketing Data Integration: CDP offerings that focus on enabling granular governance of event data streams within a marketer-friendly interface. This CDP type is used for data integration between customer endpoints and use cases around customer service and real-time-data-supported customer journeys. Example vendors: mParticle, Segment (acquired by Twilio in 2020), Tealium.
3. Smart Hub: CDP offerings that emphasize marketing orchestration and personalization. This type of CDP is a good fit for initiatives around advanced analytics, including artificial intelligence (AI), segmentation and whiteboard-style or canvas-style interfaces for customer journey. Smart hubs can also support real-time offer management with triggered messages. Example vendors: ActionIQ, Treasure Data, BlueConic, Blueshift, Dun & Bradstreet, Leadspace, Lytics, Redpoint, Simon, and Zylotech (acquired by Terminus).
4. Marketing Cloud CDP: CDP offerings that have been built upon existing enterprise marketing applications, meant to improve on the inflexible data management and profile unification features of current marketing clouds. This type of CDP is used as an evolution of the capability set that comes with existing enterprise application marketing and customer data applications. Example vendors: Adobe, Microsoft, Oracle, and Salesforce.