CDP versus DMP: What’s the difference?

This post is sponsored by Treasure Data.

Customer data platforms (CDPs) and data management platforms (DMPs) are complementary, yet different.

A CDP makes integration of first party data easier for DMPs to improve ad targeting, while a DMP enriches CDP data for smarter customer communication.

DMPs primarily focus on third party data, whereas CDPs make use of all data, including first party data. CDPs are focused on all aspects of marketing, whereas DMPs are designed specifically for advertisers and agencies to improve ad targeting.

DMPs alone won’t give you a sustainable competitive advantage over your competitors. The purpose of DMPs is to help marketers understand their audiences better so they can target their ads better.

In general, their audiences get more accurate over time because they get to collect more anonymous data as marketers use the audiences they create on the DMP to run digital advertising.

The downside is:

DMPs can’t store PII (personally identifiable information)
This is because DMPs are in the business of sharing audiences.

DMPs can’t help their customers differentiate
A DMP is inherently an equaliser, not a differentiator, and while they are excellent at complementing your customer data (so that you can find out more about who your customers are), the same information can be retrieved by your competitors and all other customers of the DMP.

CDP versus DMP

CDP features

• Use cases: All of marketing: Customer relationship management. Integrates with any take-action system, including adtech.
• Data types: First party data management with little third party data.
• Profile identifier: Primarily keyed on tangible customer attributes (PII) – customer ID, name, email, address etc.
• Data retention: Typically long retention periods to enable analytics over customer life time.

DMP features

• Use cases: Specific to advertising: Targeting ads better. Improving media buying efficiency.
• Data types: Third party data management with little (anonymised) first party data.
• Profile identifier: Primarily keyed on anonymous digital identifiers (non-PII) – cookie ID, IDFA, etc.
• Data retention: Relatively short retention periods because primary use cases are ad targeting.

To learn more about CDP versus DMP check out this report.

Five reasons to consider a CDP

1.Storing first party data, including PII
Use case: Customer identity matching.

CDP: A CDP stores all first, second and third party data, including anonymous and PII data (such as individual customer names, postal addresses, emails, phone numbers). CDPs are integrated with advertising systems, and can store DMP-type information such as cookie IDs with audience tags.

DMP: DMPs cannot accept first party data, and work almost exclusively with anonymous information such as cookies, devices, and IP addresses. Anonymity is essential to the DMP’s role as a way to exchange information about audiences without violating personal privacy.

2. Single, unified data storage
Use case: Flexible, fast querying.

CDP: CDP data is stored in a single, scalable place, making it fast and flexible to read through, and analyse all the data at once.

DMP: DMP has two different data stores – one for all the data and one for really fast utilisation of a subset of that data, but living separately.

3. Raw, detailed data with unlimited storage capacity
Use case: In-depth, rich analysis.

CDP: CDP captures raw data with granular level detail, and stores this historical information with unlimited capacity. Data is also collected for persistent, long-term storage and multiple data formats are supported, without predefined taxonomies.

DMP: DMPs collect data like common transactional marketing tools – with tags, APIs, and uploads so often the resulting view is high-level and in aggregate. As such, most DMPs and transactional tools only retain user information for less than 90 days.

4. Applications in the entire martech stack
Use case: Usage beyond advertising.

CDP: CDP can take any data and get it wherever it needs to get – regardless of where in the customer life cycle (or marketing organisation) the other systems live. They can syndicate data to any kind of partner on either side of the known/anonymous equation – adtech or martech. That can mean passing segments of rich first party data onto Facebook for look-alike modelling, or more effective re-targeting and other steps in the customer journey.

DMP: DMPs were designed to build targets for advertising and acquisition marketing, or dealing with new and unknown customers. They can be thought of as a cookie-pool with pre-built anonymous audiences to enhance display ad targeting. While they are excellent at complementing your customer data, the same information can be retrieved by your competitor.

5. Complete customer profile
Use case: Personalised customer experience.

CDP: CDP allows you to have all the information about a person – historical, contextual, demographic, behavioural, etc – on hand to inform the communication and brand interaction on a personalised basis. The persistence of the customer’s individual profile across all channels and sessions reveals a wealth of new traits and opportunities for personalisation.

DMP: DMP user profiles are basically a list of users and an associated list of the predefined categories/subcategories they belong to. A DMP builds a temporary profile based on demographic information, anonymous cookie IDs and packaged and repackaged data from one customer to another.

About Treasure Data

Treasure Data is a best-in-class customer data platform and central repository where all of your data is unified and made easily accessible into a single, reliable source of truth owned by your marketing team and available to all your business units. Visit our website to learn more about the benefits of CDP.