August 22, 2019

CDP – DMP: Are you confused ?

By Jan Teerlinck

The concept of a ‘single customer view’ is around in marketing since ever. As data is crucial, and since data management is hard, data platforms get a lot of attention. Two abbreviations go around: DMP (Data Management Platform) and CDP (Customer Data Platform). There is often confusion between the two. Let’s have a closer look at both:

The concept of a DMP originated from the adtech world, the advertisement industry. Not so long ago, display advertisement was based on context: ads for baby products are displayed on websites visited by young parents. Google and Facebook shook up this industry. Given the rich data both companies hold on us, they started selling audiences, not banner space. Seeing income decline rapidly, the traditional display industry had to come with an answer. Ten years ago, the adtech industry radically transformed. Clever tech people started building large databases with info collected from tags on many websites, enriched with all kind of sources, and connected in real-time to ad exchanges for bidding. DMPs started as third party data collection. DMPs contain anonymous information. DMPs don’t store PII data (Personal Identifiable Information). Anonymity is essential to the DMP’s role as a way to exchange information about audiences without violating personal privacy. They were designed specifically for advertisers and agencies to improve ad targeting. In the advertisement world, you don’t need to know someone's name to make a good offer. DMPs in essence are equalisers, not differentiators. Your competitor has access to the same data in the DMP. In advertisement you are not interested in what someone was looking to buy 3 months ago, you look for in-market people, the here and the now. So apart from demographics classifications, DMPs hold aggregated profile and intent data. Data freshness is more important than history. DMPs bring a lot of value for advertisers. Impossible to imagine the ads industry today without it.

Anonymity is essential to the DMP’s role as a way to exchange information about audiences without violating personal privacy. In advertisement, you don't need to know someone's name to make a good offer.

The big difference with a CDP, is that the ambition of a CDP is collecting 1st party data. Data that is collected directly from your customers and visitors, people who interacted with your company’s website, app, e-mail, in-store etc. You might say that CDP is the modern name of what used to call 'SCV’, the Single Customer View. Identity resolution is key in the creation of accurate customer profiles that marketers can use for personalized marketing communications. A CDP is designed for CRM professionals. CDPs can contain both identified and anonymous contacts. Think of website visitors that have never registered, same for your app users. A key requirement for a CDP is that it connects to your marketing systems. Because eventually you store customer profiles for creating highly personalized interactions with customers.

Let’s line up some main differences:

Data type


Almost exclusively anonymous.
Profiles are unified by things such as cookies, IP addresses, device ID’s.
Mostly 3th party collected from multiple sources.


Mostly identified PII information.
Storing all information about contacts that interact with you (customers and visitors).
Designed for 1st party data.



Optimize advertisement. Create targeted audiences. The focus is on segmenting and categorizing anonymous people


CRM: Central collection of all customer data, with the purpose to feed that to all your execution systems for personalized interaction.



Data freshness is more important than history. Data is often stored according to pre-defined taxonomies. Retention times of 100 days or less.


Data is collected for persistent, long-term storage. Formats are free. While in theory designed for unlimited capacity, that doesn’t exist. Retention time is free to choose by marketing based on relevance.

As always in martech, evolution goes fast, systems converge, and that adds to confusion. For example: Some CDP’s are integrating with advertising systems. That makes totally sense and is what you want. As a result they are storing more DMP-type information such as cookie IDs with audience tags. Other tools, also naming themselves DMP, focus on 1st party behavioral data, allowing to ingress your CRM data. These are carefully isolated from situations where anonymity is still important.

The lines are blurring, but the essential difference remains. Convergence of systems and functions is a constant in martech. Don't let that delay your marketing efforts and investments. But how does your campaign management system fit into this? How do those vendors look at the concepts of a CDP and DMP? Do you want again another data store? I know this conversation is ongoing in many companies. There is a lot to say on this topic. That is for another blog post. In the mean time, reach out if this plays at your company. Happy to help.

Share this
Is this helpful?
If this topic plays for you and you want to dive deeper into it, send Jan an email at