As Advertisers enter the process of determining whether they need a Data Management Platform (DMP) - or, having already decided they need one decide which DMP they need - the emergence of a new technology known as Customer Data Platforms (CDP) can often confuse things and muddy the waters.

How can the two platforms be distinguished and where does their functionality overlap?   

Often people differentiate between them by saying that CDPs do what DMPs do with the added benefit of being able to use personally identifiable information. So essentially, using a CDP, Advertisers can gather and segment personal data from multiple sources, configure it, and make it available to marketers for activation in their digital marketing efforts based on a particular customer’s history of touch points. If a customer - let’s call him Sam - opened a particular marketing email, registered for a webinar, or downloaded a white paper, then Advertisers can tailor a marketing story for Sam based on his interactions with the brand, or, in other words, on his “first-party data” points.

Sounds great, but how different is this from a classic DMP use case? 

Primarily, it depends on the way the DMP is integrated into the advertiser’s technical infrastructure. If the DMP just gathers touch points on the advertisers’ sites or shops to allow segmentation and categorization based on relatively vague online contacts, then a CDP may seem like a better option. However, this is not the way a DMP should be implemented if you want it to fulfill its purpose. 

The biggest misconception around DMPs that I have heard in the market is that they focus on “third-party data” – which couldn’t be more wrong. A DMP is, in fact, all about “first-party data”. Every advertiser should aim to integrate his DMP into their existing technical infrastructure. Because, if a DMP is integrated into an advertiser’s CRM or general business intelligence systems then the only difference is the fact that Sam is no longer Sam but a privacy compliant, pseudonymized ID instead - and that’s about it.  

Of course, there are other important distinctions beyond my simplified example above. For example, a CDP always works on deterministic user information instead of probabilistically stitched users. And a CDP does have a valid long–history of user interactions which are accurate because of the named deterministic matching, which automatically enables functionality across devices.  

However, again, is cross-device purely the reserve of CDPs? Absolutely not. If the pseudonymized DMP IDs are matched to the persistent IDs within the advertiser CRM system, all the above-mentioned functionalities apply to the DMP as well. And they apply without exposing personally identifiable information to an external network, which is a substantial consideration in the days of GDPR and ePrivacy.

Companies have crunched their data internally for years now. Many have learned a lot about their customers and have applied these lessons. Why reinvent the wheel with a brand new external system? Instead, allow your existing data to reach out into the online world via a DMP. I’m not saying that CDPs are a terrible thing, in my opinion, the differences between a DMP and a CDP often overlap, and advertisers have to consider what they already have available internally when making a decision.

If you have CRM data that you want to activate by integrating into the Adform Stack, feel free to contact your local account manager today.