Have you ever wondered what’s the difference between the various types of systems that store customer data?  Of course not – you’re an industry professional!  But maybe you have a friend who isn’t quite so well-informed.  Here’s a quick list you can share to save yourself the work of explaining it to them.  The definitions (in italics) come from the CDP Institute’s Glossary of CDP Related Terms, which you can read here or download here.

Definition of Customer Data Platform (CDP).  The CDP Institute defines CDP as packaged software that builds a persistent, unified customer database accessible to other systems.  In other words, a CDP is a specialized system designed to assemble and share customer profiles.  Many people would extend the definition to include features that apply the customer profiles to business purposes such as analyzing data or orchestrating customer interactions.  But while many (in fact, most) CDPs do provide such capabilities, we don’t consider them essential to earning the CDP label.  (For a more detailed definition and criteria, see the CDP Institute’s RealCDP program.)

Taking the CDP definition as a reference point, let’s look at some other systems that also manage customer data.

Definition of Master Data Management (MDM).  Software that reconciles different versions of information about an entity (person, product, location, etc.), selects the version to be used as a standard across company systems, and shares this version (called a “golden record”) with those systems. MDM systems may perform identity matching as part of their function.  Traditional MDM systems only store information about entities, although modern MDM may store additional information such as transactions related to an entity.  CDPs that lack their own identity resolution capability may rely on an MDM for this.  Similarly, a CDP may rely on data cleaning and standardization provided by an MDM if that is available.  Compared with a CDP, MDM systems have a broader scope (i.e., not just customer data) but less depth (i.e., lacking behavioral details), and are not designed to support customer analytics or data activation.

Definition of Customer Relationship Management (CRM).  Software that stores details of direct interactions between a company’s customers and its sales and service personnel.  Many companies treat their CRM as the primary customer data store, especially in business-to-business marketing.  They may even refer to their CRM as a CDP.  But, as the definition indicates, the CRM is largely limited to data about interactions that are executed within the CRM: most CRMs have little or no ability to import data that originates in other systems or to do the identity matching necessary to merge external data into comprehensive customer profiles.  Furthermore, the transaction-oriented data architecture of a CRM is geared to working with one customer at a time; as such, it is poorly suited to the bulk processing needed to analyze or activate the customer database as a whole.  In general, the CRM would be a data source and data consumer for the CDP, not a replacement.

Definition of Data Management Platform (DMP).  Software that stores anonymous customer profiles, primarily to support Web display advertising.  Traditional DMPs were designed to manage cookies, which were segmented and selected as advertising audiences.  As privacy constraints reduced the availability of cookies, this became less important and many DMPs either went out of business or evolved to add other, CDP-like capabilities related to first-party data and known customers.   Some have actually repositioned as CDPs, although not all those firms have re-engineered their products to support all CDP use cases effectively.  Today, marketers are more likely to ask their CDP to provide DMP functions than to ask a DMP to double as a CDP.  To meet this demand, many CDP vendors have added support for advertising use cases and connectors to advertising systems, something that was not common in earlier times.

Definition of Marketing Automation Platform (MAP).  Software that maintains customer and prospect lists and runs campaigns against them. Primarily used for outbound campaigns (e.g., email) but some systems also support real time interactions (e.g., Web site messages). Largely limited to data generated within the system itself and to imports from CRM systems.  Early marketing automation systems, such as Eloqua and Marketo, primarily supported business marketers and integrated closely with CRM systems.  Over time, the term expanded to include “campaign management” systems for consumer marketers as well.   Although marketing automation systems were always designed to ingest external data and build unified profiles, they were generally not designed to share the resulting profiles with other systems.  Many marketing automation vendors have subsequently removed this limitation, enabling their systems to qualify as a true CDP.  As previously noted, the majority of CDPs today provide some set of marketing automation functions, either because they originated as marketing automation products or because they added those features to a basic CDP.

Definition of Privacy Management Platform (PMP).  Software that helps organizations to manage personal data, including data inventory, consent management, compliance, records of processing activities, risk and privacy impact assessments, and data subject request processing.  Unlike CRM, DMP, or MAP systems, privacy management platforms are not major repositories for customer data.  Rather, they are tools used by privacy managers to track where customer data resides and how it is used.  This means they are complementary to CDPs but not potential substitutes.  Some CDPs include some privacy management capabilities, although most CDP vendors have chosen to integrate with third-party privacy systems rather than encompass them.

Definition of Data Warehouse (DWH).  A collection of data copied from company systems, reorganized and often summarized for analysis. The data warehouse concept predates CDPs by several decades.  Like CDPs, they were designed to provide collect data that originated in other company systems but was not easily accessible.  Original data warehouses were used primarily for financial analysis and employed specialized data cleaning processes and data structures tailored to this purpose.  Most were not designed to store customer-level details or unstructured data, or to allow direct access for activation tasks such as audience selection or real-time queries.  In fact, CDPs evolved largely because these limits meant the data warehouses available at the time could not meet marketers’ requirements.  More recently, “cloud data warehouses” such as Snowflake and Amazon Redshift have removed some of these limits, although others remain.  Today, some CDPs use the same database technology as cloud data warehouses, although they often supplement this with special features to compensate for the cloud data warehouses’ limitations.

Definition of Data Lake.  A collection of data copied from company systems, stored in its original forms and accessible for analysis and further processing. The first step in building a data warehouse is typically copying data from source systems in its original format into staging tables; the data is then cleaned, standardized, and otherwise transformed before loading into the warehouse itself.  These staging tables have now taken on an identity of their own as a data lake.   Because they retain the original detail of the source data, data lakes often are a more useful source for a CDP or CDP-style customer profiles than the summarized data in a data warehouse.  However, the raw, siloed nature of the data lake files means they require additional processing to be transformed into the unified customer profiles that support CDP use cases.  While some of this processing can be done on-demand by reading the data lake files directly, it is usually more efficient to do most of it in advance and store the results in a warehouse or CDP.

Definition of Composable CDP.  An architecture that uses separate components to deliver CDP functionality, often based on a data warehouse built separately.  Strictly speaking, there is no such thing as a composable CDP system; rather, composability is an architecture which includes systems as components.  Most composable CDP vendors started selling a single component, usually to access data in a warehouse (“reverse ETL”).  These vendors have nearly all now expanded their product lines to include several components, with the goal of offering a complete set of CDP functions.  Despite this expansion, they are still designed to work with an external data warehouse as the primary store for customer data.  These vendors increasingly refer to themselves as “warehouse native” rather than “composable CDPs.”  At the same time, many traditional CDP vendors have split their products into components that can also be installed separately and attached to an external data warehouse.  The traditional CDPs also increasingly offer options to combine data stored in an external warehouse with data in a CDP database.  As a result of these trends, the distinction between traditional and composable CDP vendors is increasingly vague, even though the distinction between a warehouse-native and CDP-based architecture remains clear.

Definition of Customer Data Infrastructure (CDI).  Technology that manages customer data, including software, hardware, and business processes.  Core functions include data governance, data quality, identity management, profile creation, and privacy compliance.  As the definition states, infrastructure is a combination of systems, hardware, and processes.  Vendors that refer to themselves as CDI systems generally provide the functions to build and maintain a customer database.  This is roughly the same as the core profile-building functions offered by a CDP, although it may extend to supporting capabilities such as data governance and quality management.

Understanding Customer Data Management Systems

We hope this provided a good primer into customer data systems so you can understand how they might interact and to help you select the right technology for your company. To learn more about CDPs and different topics around customer data, see our library of resources here: CDPI Library.