What is a Data Management Platform?

Marketers want to make their connections and interactions as relevant as possible to their target consumer. Therefore, a marketing message targeted to an audience is more relevant than a one-size-fits-all message. Targeted messages help marketers improve on customer acquisition and retention.

In the online world, marketers have the potential to understand and communicate with ‘like-minded’ groups of individuals in targeted ways through Data Management Platforms (DMP). DMPs help brands have a consolidated view of their audience, and help them create marketable segments based on audience characteristics and behaviors. Leveraging these segments enables the brand serve targeted, more relevant content to their audience.

Specifically in the advertising world, a DMP also helps publishers understand the types of audiences visiting their properties and channels and use that to power content engagement strategies, advertising programs, customer acquisition and retention programs.

The components of a DMP

At its most basic level, a DMP does determines two things in order to bring marketers a step closer to better target the right person with the right marketing message:

• Identity: In order to target an audience, a DMP needs to be able to first determine who that audience is in terms of their identity, regardless of the channel or device they are coming form.  And it needs to do that without storing any PII.
• Behavior: The second part is to determine what that audience does in terms of their behavior, which includes audience characteristics, context and actions.

A DMP has four key capabilities :

1. Data Collection: A DMP consolidates all of the different data sources it has about a user (first party data) and matches the data to create a single view.
2. Identification & Data Enrichment:  A DMP attempts to uniquely identify an individual consumer across devices and channels and links together all data about that consumer. Identification of a consumer helps it enrich the data records with additional first, second and third party data.
3. Segment Creation: With all the data in a single view, DMP discover audiences and build audience segments, based on rules created manually or algorithmically. The rules are driven by audience characteristics, contextual and behavioral data.
4. Segment Delivery : The segments are delivered to a brand’s digital channels such as their website or offsite campaigns such as email or display Ads. The platforms running these campaigns leverage these segments to target their messages personalize the experiences.

In order for the DMP to be able to do all of the above, one of the biggest components of a DMP are the integrations it has with other platforms. These integrations comprise of data integrations such as Analytics, 3rd party data providers, Tag Management platforms, Attribution platforms, consumer data inventory sources; as well as targeting integrations with web CMS, Ad Servers, DSPs etc.

Our Approach

In this post we will take the example of a hypothetical brand called fitness.at.work to illustrate the components and capabilities of a DMP.

Below is the journey of a fitness.at.work customer. We will see the role the DMP plays in this journey.

Data Collection

A DMP helps marketers collect first, second and third party data.

1^st Party Data: This is data that a brand collects from direct interactions with the consumer, which could include online browsing behavior on the brand’s sites, purchase transactions, social data that the consumer granted the brand access to, etc. As the DMP collects data it starts to create IDs to uniquely identify individual users.

In our example customer journey, the first party data is collected from the online webstore as well as the physical store.

2^nd Party Data: This is data from a brand’s partners who have acquired data about the consumer, which its partners decided to share with the brand through some mutually beneficial agreements. This data is added to the 1^st party data for an ID.

In our example for fitness.at.work,  second party data could be information that’s voluntarily shared by their client company about their employees. This could be data such as average work hours, general wellness goals, etc. at an aggregate level.

3^rd Party Data: These are generic data points such as demographic, psychographic, lifestyle data, etc. of people in a brand’s their target market. This data is collected from data providers and data exchanges and aded to the 1^st party data for an ID. DMPs are usually able to integrate with partner data providers outside your organization through real-time or offline mechanisms. This enables your brand get the benefit of the third party data synchronized with the DMP.

Many brands work with third-party data match partners such as datalogix, axciom, etc. to enhance their customer data. DMPs can usually integrate with these data match partners and sync a brand’s audience information with those matched with the data match partner.

Tag Management is a must-have integration to enable a scalable method of data collection into a DMP. JavaScript tags are used in websites to collect the data into the DMP. Since, marketers want to tag many components and interaction points on a website to get as much visibility into user behavior as possible, there is a need to effectively create, manage and deploy these tags, which can be accomplished with a Tag Management Platform that works in tandem with the DMP.

Matching

Segment Creation

Within the DMP, the 1^st, 2^nd and 3^rd party data sources are made accessible to marketers to enable them to define rules based on that data. And these rules are combined into segments. An example rule within our hypothetical business fitness.at.work could be:

Rule1: c_prop1 equals ‘fitness gear’

Here c_prop1 is the data point that’s sent from the Tag Management System into the DMP when a person clicks on a product listing page (PLP) for fitness gear. The value within c_prop1 could be set to the name of the PLP that the user clicked. Therefore, when a loads the fitness gear page, it triggers Rule 1. These rules are also called traits and all the traits within a DMP represent a trait taxonomy. Multiple traits are combined into segments.

So if you had another rule

Rule 2: c_prop2 equals ‘Women’

Then, you could define a segment of ‘in-market for female fitness gear’ as

in-market for female fitness gear if: Rule1 is true AND Rule2 is true

A user is assigned to this segment if they satisfy both Rule1 and Rule2 based on their navigation on the site.

The process of assigning a user to an audience segment is called qualification. The qualification is done in two ways:

Real-time qualification: This happens when a user matches the rules within the segment.
Batch qualification: This happens when new segments are defined or when segments are modified. The matching of segments to data is then retroactively qualified to audience data collected in the DMP.

Upon qualification, the DMP is able to display the segment population within ‘in-market for female fitness gear’ to the marketer.

Segment Publishing

DMP segments are published to DMP destinations.

A DMP destination is any third-party system (an Ad server, Website, Email marketing tool, DSP, etc.) is where the segments of the DMP are shared with. Sharing a segment tells the third-party system which segment a particular user belongs to, and the third-party system can utlilze that intelligence to target the marketing message.

As an example, based on the segment of a user, the homepage of a website could display a specific promotion. Or an email marketing tool could use some special text or offer within an email to a user belonging to a segment. And finally, one of the biggest use cases for a DMP is for sharing segments with DSPs in the programmatic bidding process, so that an Ad is served based on a matching segment.

Sharing of segments works because of ID syncing between the DMP and thid-party systems.
Since data transfer mechanisms vary across third-party systems, the DMP usually allows to standardize data transfer mechanisms

Performance Measurement

Performance measurement is done through the reporting capabilities of a DMP. There are usually 3 types of performance reports:

Taking action based on segments

Segments are used across marketing channels (called targeting destinations) such as Ad servers, content optimization, content personalization, creative optimization, video delivery, search, etc. A destination is a third-party system (e.g. An Ad network) that a brand wants to share data with