On May 3, in celebration of the newly coined International MarTech Day, Scott Brinker released his MarTech Technology Landscape supergraphic with 9,932 vendors in 49 categories, representing a 22 percent growth since the release of the previous graphic in 2020. In an attempt to cut down on the infamous complexity, some organizations make the mistake of thinking that differently named applications are the same thing. Case in point is a master data management solution (MDM) and a customer data platform (CDP).
We have the first, so why would we need the second is a common question from companies looking to streamline their martech stack. While there is some overlap, the short answer to the question is that an MDM and CDP have substantively different goals, target audiences and core use cases. We’ll explore the key similarities and differences and explain why MDM vs. CDP is not a straightforward either-or decision.
MDM and CDP View of Data
An MDM, typically an IT initiative, is a solution whose goals encompass mastering the data realms for all the entities that make up an enterprise; not just the customer or party entity, but a range of domains that include product, site, contracts and service. These domains are inherently related to what a CDP does, with a key difference that the CDP’s focus is on customer data. For instance, a CDP records transaction information which by definition includes product information, but a CDP is not trying to perfect the product, it is only trying to accurately relate the product to the customer. The same distinction holds true with the other domains; a CDP may require accurate site information to drive an offer when a customer breaks a geofence, but mastery of the site data is not within its purview.
We can also think of the key distinctions in reverse. That is, an MDM has an interest in perfect customer data so that IT applications related to governance, supply chain, CRM and service and cost management have accurate customer data, while a CDP’s interest in in perfect customer data is to create a perfect customer experience.
The audience for a CDP, then, is not IT but marketing, which means they are using a CDP a source of truth about a customer to execute against use cases such as acquisition, engagement, retention and upsell. The different focus or goal of the product creates interesting consequences. Even in the customer domain, an MDM will care about core customer attributes such as having the correct name and address, and having that accurate information as it relates to other customers. But an MDM is not going to care about the customer’s behavior or other attributes that will help analyze customer intent.
MDM and CDP Areas of Overlap
While both a CDP an MDM will have an interest in creating a golden record, the former will use it for the sole purpose of improving customer experience, while the latter will use the unified customer record to standardize service engagements or enterprise governance applications, for instance.
Looking at specific data areas of interest, order and shipping details is one area that highlights the distinction. A CDP will be interested in the information to resolve a customer inquiry perhaps, but for an MDM the information is critical for enterprise resource planning (ERP) and supply chain management. Similarly, semantic definitions and constraints, such as defining a master key for various entities, standardizing currencies and measurements, etc., are typically owned by an MDM while a CDP will inherit those definitions and constraints. Likewise, because MDM owns master data, any time there is a change to a description of a product, any system – a CDP included – that carries a description of the product will need to inherit the change using the MDM as the source of truth.
To summarize the overlap, the primary areas are:
- Building a unified profile, but with different data and goals (MDM for domain app standardization, CDP for improving CX)
- Identity resolution for relevance and compliance (MDM for details of semantics and governance, CDP for CX in marketing)
- Data quality for reducing errors and ambiguity (MDM for enterprise control and unity, CDM for details of interactions/behaviors)
See Figure 1 for side-by-side comparison details.
|• Operational Systems (primarily CRM)
• Transactions (online & offline)
• Martech (behavior, offers, responses)
• Predictions & Aggregations
|• Operational Systems (ERP, CRM, etc.)
• Transactions (including PIM)
• Operational “best value” selections
|Data Governance & Stewardship
|• Basic assessment of accuracy
• Can include stewardship for privacy (GDPR, CCPA, etc.)
|• Data Modeling
• Metadata repository
• Entity definition & resolution
• Data constraints and relations
• Stewardship for detailed control
• Enterprise distribution or availability with access management
|Data Cleansing and Quality
|• Varies… can be very simple parse & cleanse or rich normalization and cleanse of name, address, phone with profiling and trending of quality
|• Generally, follows enterprise-wide rules for cleanse, normalize, enrich, with profiling and constraint / anomaly rules
|• “Profile Unification”
• Usually, simple deterministic rules
• Can use probabilistic or ML-based rules
• Relation to “household” or “organization”
• Specific to use case / purpose
|• “Entity Resolution” – Party, Product, etc.
• Complex rules with human oversight
• Rules specific to industry and domain
• Relation to all covered entities
• Enterprise-wide fixed rules
|• Core capability of CDPs to support marketing use cases and activation
• Analytics for assessment, optimization
|• Generally absent
• Analytics for metadata and stewardship
|Publish segments into martech and ad tech “channels”
|Generally absent, but see Stewardship
A closer look at the comparison details, specifically data stewardship and governance, reveals why an organization with an MDM may also need a CDP. If for instance data stewardship capabilities are needed for entity definitions and resolution, data constraints and other reasons listed in Figure 1, a company should examine whether its current tech stack has those capabilities and, if not, determine if an enterprise-wide approach using MDM is worth the investment or whether a domain-specific approach will suffice. Because if the purpose of data governance is solely to handle data subject requests under GDPR, perhaps a specific privacy and consent solution integrated to a CDP will meet these goals better than an MDM would.
Applying that same calculation for data scope, identity resolution, segmentation and other key purposes will help an organization make the right determination. As we’ve seen, there is room for both an MDM and CDP depending on an organization’s reasons for collecting, storing and using enterprise data, and more specifically customer data. Broad enterprise goals will tilt toward MDM, where a strict focus on improving CX will tilt more heavily to a CDP.