Data management platform (DMP)

What is a DMP?

A data management platform (DMP) is a marketing tool that manages incoming customer-related, demographic, mobile, and cookie information. As that data comes in, the DMP gathers it and uses the processed data to create anonymous customer profiles for various audience segments. An organization may build its own DMP tool or partner with an existing software provider, such as Nielsen and Oracle.

Types of data DMPs use

There are three broad categories of data that a DMP uses:

• First-party data: Data gathered and owned by a company, including website or mobile app data or customer relationship management (CRM) platform data.
• Second-party data: Data provided by another company the organization has a mutual agreement with, including customer journey or online campaign data from a partner or vendor.  
• Third-party data: Data purchased or obtained from social media platforms and other analytics companies, used to supplement first-party data. This can include data collected through website cookies.

DMP vs. CDP

DMPs and customer data platforms (CDP) are marketing tools that can work together to help marketing teams. The main differences are the types of data that they work with, how long they store that data, and how they use the data. 

DMPs

• Primarily uses second- and third-party data.  
• Gather data from multiple sources about multiple customers to generate a general audience profile, so that marketers can target advertisements to certain types of customers. 
• Provides general, anonymous profiles and does not provide specific information about individual customers.
• Retains data for short periods before converting it into an anonymous form.

CDPs

• Primarily uses first-party data but can also interact with second-party and third-party data.
• Stores and organizes information about a potential customer from multiple sources. 
• Acts as a centralized location for customer data, including product, marketing, sales, and more. 
• Builds a more individualized customer account based on the website they came from, articles or blog posts they read, purchases they make, product usage, and other interactions.
• Keeps data in its original form for long-term use. This period may vary depending on the solution, but it is an important distinction because data retention policies can have implications for privacy laws.

To generate a general marketing persona to target a type of customer on a paid ad platform, consult the data in your DMP. To send a personalized marketing email to your 100 highest-spending customers, consult the CDP. 

Do I need a DMP?

To determine whether you need a DMP, consider whether its benefits and uses will save the team time or money, improve conversions, or improve the marketing processes the team already has in place.

Benefits of DMPs for marketers

A key benefit of using a DMP is reaching a wider audience. By augmenting the organization’s customer insights from first- and second-party data with third-party data sets, they can reach new customer groups. 

A DMP can also centralize and manage the data for multiple marketing campaigns at once, analyze information about the customers that it reaches, and monitor users across multiple types of devices. Some tools include encryption capabilities, which can help protect data in case of a cybersecurity incident.

Marketing use cases for DMPs

Take, for example, a product that’s popular with a segment of 20- to 30-year-olds from the Midwest US. The marketing team likely already has a list of customers and prospects who know about or work with the company. A DMP can help create a broader list of people to market toward on a paid ad network. Not only people in that age range or physical location, but also look-alike prospects in other segments. This can help the organization personalize advertising campaigns, make better use of the advertising budget, and reach new customers.  

Challenges of DMPs

As a user visits a website, their browser stores data from the website as cookies. Because DMPs rely on cookies, any browsers or browser plugins that block cookies can impede their functionality. 

Under the European Union’s General Data Protection Regulation (GDPR), organizations can only use third-party data collected through cookies and other internet activity if they have user consent. While a DMP uses anonymous data — and typically does not store personally identifiable information (PII) — GDPR applies to all businesses that interact with any European customers Chapter 3), which is why many websites now prompt users for cookie consent by default. 

Additionally, Apple, Google, and Mozilla, makers of three of the most popular web browsers, have announced plans to phase out third-party cookies in upcoming versions of their browsers to protect user privacy.

How does a data management platform work?

To set up a DMP, you first connect it to the data sources it will work from, such as a CRM, website, or social media. If it collects any first- or second-party data, the DMP will take in that data and combine it with third-party data sources.

Using internal algorithms, business intelligence (BI) metrics, and in some cases, artificial intelligence (AI), the DMP uses the data to create audience profiles. For example, it may recognize an existing customer from their browser ID and prompt the company website to serve them a relevant ad. Then, it can use an associate ad exchange to serve similar ads when the user visits other websites. 

The DMP may simultaneously serve the same ad to other users with similar demographic details, such as age, location, or racial background. 

About the data pipeline and DMP architecture

The “data pipeline” is a data management concept first proposed in the late ‘90s, which refers to a series of data processing tools that work in a sequence to collect, process, and store data for other uses. 

• Collection: The DMP gathers raw data from a variety of sources. In the case of a DMP, this would be first-, second-, and third-party data from a company website, partner companies’ tools, and any other sources the DMP can integrate with. 
• Transformation: The DMP automatically processes the data from its original format into a usable format for marketing or advertising channels. For example, a DMP may take data entered into a web form and convert the unique, customer-specific data into anonymized demographic data.
• Storage: The processed data goes into the DMP’s data repository, where it can be used in the DMP’s other functionalities, like generating profiles for look-alike customers. 

“Data architecture” refers to the methods and rules governing how data is structured and organized within a data repository. For a third-party DMP tool, much of this work is done by the DMP itself. Working with the DMP vendor in the early days of using the tool helps ensure the data gets into the DMP’s pipeline correctly. 

For an organization planning to build its own DMP tool, building the data pipeline and architecture will be a more complex process. It requires the team to consider data access and sharing, implement security measures for the data, establish a common vocabulary for what data points mean, and determine how to minimize how much the data gets altered during processing. 

Features of a DMP

When speaking with a potential vendor, ask about the following key features of the DMP:

• Advertising data integration: DMPs can handle a variety of data from multiple sources, so talk with the vendor about the sources it can ingest and work with. 
• Audience analysis: A DMP should provide information about how well an ad campaign did before and after it ran, so ask about the insights it can provide.
• Audience building: A key feature of a DMP is building audiences based on customer and prospect data, so ask how the tool enhances targeting data and exporting audience profiles to other platforms.
• Cross-device targeting: As potential customers already use multiple devices, ask how the DMP targets users across phones, computers, tablets, and smart TVs.
• Security configurations: Research and ask about the encryption settings, security controls, and how to configure the DMP for the strongest security posture.

How do I implement a DMP at my organization?

Before you implement a DMP, develop a clear plan for your implementation process. In addition to the normal steps for acquiring and implementing a new tool, and any change management processes, focus on developing your goals, strategy, and processes for the data, and working with the vendor.

Develop goals and a data strategy

Simply buying a DMP tool will not change a marketing program overnight. Decide beforehand what the tool should accomplish. Work with the marketing and communications teams to identify which features and benefits the DMP must provide, and how it could strengthen various campaigns. Set specific, reasonable, and actionable goals with metrics for success with the data management platform solution.

Consult any data or analytics teams that work with the data to get a picture of how that data enters data warehouses, how it is used, and what data is available. These teams typically already have a data strategy, and can help develop one for the DMP.

Create a process for using the data

While some of a DMP’s functions are automated, someone needs to work with the tool. When implementing a new solution, any documentation about that tool gives the users a starting point. The more processes written and recorded before and shortly after implementing the tool, the easier it is for people to use the DMP. User guides and cheat sheets can always be updated as you gain a greater understanding of how to use it. 

Collaborate with your DMP vendor

When purchasing a DMP from a vendor, they will likely have a customer success team dedicated to making sure their customers’ needs are met. Before choosing a DMP solution from a vendor, talk with them about your goals, and ask them about how to meet the established success criteria. Ask about reports or statistics the tool can provide for metrics related to these goals. Keep the vendor’s contact information handy and reach out with any questions. 

Continually measure progress against goals

Setting goals and metrics early will help measure the program’s success. Check in with these goals on a regular (such as monthly or biweekly) cadence, comparing the metrics collected from the DMP against the established, measurable targets. There may be a learning curve with a new tool, and some goals may start slower than others. This may not mean that the DMP is not working, and may indicate that the process requires adjustments or patience. 

How do I choose the right DMP for my team?

Not all DMPs are the same, although many have similar features. Each vendor solution will work in slightly different ways, so make sure the marketing team’s needs meet the realities of your technology stack.

Who should be involved?

As the tool touches multiple roles, implementing a DMP should not be left to one or two stakeholders. Depending on the organization’s structure, seek input from the following groups:

• Marketing team. This team will be working with the output from the DMP most closely. Consulting team members most affected by the DMP will ensure that the tool helps everyone.
• Someone who works with data collection or data ingestion points. Consult the analytics and data teams to understand how data is currently structured and ensure the DMP can work with it.
• Any vendors the DMP will connect with. Some DMPs may integrate with the tools the organization already has, and some data may need additional work to be compatible. Speaking with vendors helps to understand what integration features the tool needs.
• Your financial or budget director. Whether building or purchasing a DMP tool, consulting the finance team clearly shows how much money is available.

Reading DMP reviews

Many industry publications and review websites include reviews for DMPs, such as Gartner and Slashdot. When it comes to reviews, remember that everyone’s situations are different, and a tool that does or doesn’t work for another company may have different results for you. Read the full review to help compile a list of questions for the vendor, and compare the reviews to the goals for the DMP.

Data management platform companies

While researching your options, the following platforms may be good places to start:

(presented in alphabetical order) 

• Adobe
• Google Audience Center
• Lotame
• Nielsen
• Oracle
• Permutive
• Simplifi.fi