David M. Raab is the Founder and CEO of the Customer Data Platform Institute. A recognized marketing technology expert, he named the Customer Data Platform (CDP) category in 2013. After starting his career as a marketer, he spent years as a technology consultant to Global 2000 firms across retail, media, communications, hospitality, and technology industries, helping businesses optimize their customer data strategies to enhance engagement and marketing performance.
Raab has written hundreds of articles, significantly contributing to the advancement of data-driven marketing, and is the author of The Marketing Performance Measurement Toolkit. He is a graduate of Columbia University and the Harvard Business School and has presented at industry conferences worldwide.
In this exclusive interview, we explore the origins of CDPs, their influence on CRM, common challenges in CDP implementation, and the future of AI-driven customer data strategies.
What is a CDP
A customer data platform (CDP) helps you organize and consolidate customer data from various sources. Be it data from in-store and online purchases, website browsing behavior, app usage, or phone complaints, a CDP standardizes this data and brings it together in one central platform, eliminating data silos.
The result is a complete, 360-degree view of the customer, allowing you to develop personalized marketing campaigns precisely tailored to the needs of your target audience.
Insights on Customer Data and AI
Q: You famously named the Customer Data Platform category in 2013. What gap in the market prompted you to recognize the need for a CDP?
David: At that time, I noticed software applications, such as marketing automation and predictive modeling systems, building their own customer database.
This was unusual at the time because the traditional approach had been to create a separate, custom-built customer database and attach the application to that. This enabled one piece of application software to work for many different users, but it also meant that software could only be sold to companies that had a customer database in place. Building their own database enabled the applications to serve additional companies.
In any event, the idea of a packaged rather than custom-built customer database was new. The concept needed a name, and the letters ‘CDP’ were available as an acronym, so that’s what I chose.
Most CRMs lack the data management capabilities to integrate online and offline data, and they are not designed to send data to advertising platforms or connect to a BI tool.
On the other hand, CDPs can accomplish this by combining data from various online and offline sources (e.g., physical purchases) in one place and using them for in-depth analytics and tailored marketing.
Q: How have CDPs reshaped (or are reshaping) traditional CRM strategies over the past five years?
David: In B2B, CRM is often considered the primary customer data store. This has slowed the adoption of CDPs in that space because the need was not so clear. In B2C, CRM systems are less central, so they are not as easily confused with CDP.
More recently the distinction between CRM and CDP is better understood. In particular, CDP captures web behaviors that a traditional CRM does not. It also captures other data types that are usually not part of CRM, such as purchase history. As the need to develop customer management strategies that incorporate data from non-CRM sources becomes clearer, the role of CDP becomes more central, and CRM returns to its original purpose of managing sales and service interactions.
The CDP also enhances these CRMs by providing them with more complete profiles than the CRM can assemble since this is limited to data generated within the CRM system.
"As the need to develop customer management strategies that incorporate data from non-CRM sources becomes clearer, the role of CDP becomes more central."
Q: What is the most frequent mistake companies make when deploying a CDP?
David: The most frequent mistake is failing to engage with users to define target use cases during the selection and planning stages. This has several bad consequences, including selecting a system that doesn’t meet their requirements, users who don’t know what the system is for, and a lack of cooperation among different departments.
The second-most frequent mistake is buying a system based on price rather than features. This also traces back to a lack of user engagement since companies buy on price when they think all competitors’ features are equivalent. It’s only the users who understand their needs in enough detail to recognize which subtle feature differences are important.
Q: How should businesses align their CDP and CRM strategies with data privacy and compliance requirements?
David: Privacy and compliance need to be built into their customer data strategies. The CRM is a logical place to capture and store consent, while the CDP is the logical place to apply consent and other compliance constraints when using customer data.
Having the CDP as the primary source of customer data for all activities makes it easier to ensure those activities are compliant and to keep the data use records required by privacy regulations.
That said, privacy management is a complex task, so many firms that keep customer data in a CDP will integrate it with a separate Privacy Management System to handle the processing details.
New and stricter data protection laws, such as the General Data Protection Regulation (GDPR), Loi Informatique et Libertés (LIL), and the California Consumer Privacy Act (CCPA), make CDPs indispensable for managing customer data compliantly and avoiding significant penalties. CDPs provide a secure environment for storing and processing first-party data for better documentation and control of data usage. This reduces the risk of data breaches and strengthens your customers' trust in how their data is handled.
Q: AI is increasingly touted as a game-changer for data-driven marketing. How do you see AI intersecting with CDPs in the next decade?
David: So far, we’ve seen AI enhancing existing CDP functions such as source data modeling, quality management, identity resolution, entity extraction, and analytics (segmentation, predictive models, etc.)
At some point, AI may take a more central role in the CDP, automatically accessing the right data for particular purposes even if it isn’t already loaded into the CDP. This would be closely related to using AI to design marketing campaigns and generate marketing content automatically.
At the moment, AI for such tasks relies heavily on CDP data but has yet to create that data itself. If this becomes the case, it could replace the traditional CDP since they would organically generate customer profiles as needed rather than reading pre-existing data from the CDP. The challenge with this approach is ensuring that you don’t lose other benefits of a central, stable CDP, such as privacy and consistent treatments across channels and over time.
Q: What are your predictions for CDPs and customer engagement?
David: It’s pretty clear that the current trend is for CDPs to be embedded within customer engagement systems. This helps the customer engagement vendors migrate from the application layer –where they can be replaced by other, similar applications relatively easily – to the platform layer, where making a change is much harder because the platform is connected to so many other systems.
The problem is that companies have many applications, and they can’t all migrate to the platform layer since the whole point of a platform is you have just one. So, while you can see why application vendors add platform capabilities, it’s not necessarily in the users’ interest to adopt them.
The smart vendors will accommodate this by ensuring that using their built-in CDP is optional, so a buyer can use their application connected to a different CDP if that’s their preference. The built-in CDPs will connect with other vendors’ applications for the same reason. While nothing is guaranteed, the net result should be a greater choice for users.
Q: How might “agentic” or autonomous CRM tools—capable of making proactive, data-driven decisions—fit into your vision of the future landscape?
David: Autonomous CRM tools are revolutionary in many ways, but they don’t change the architecture of your customer systems. Whether customers interact with a human agent or an AI agent, the same data goes into the CRM, and the CRM has the same relationship with the CDP and other systems.
At most, you have a change in degree, where the autonomous CRM can exploit more data more quickly, and other systems must meet that increased demand.
It’s possible that agentic systems will be inherently multi-channel, which might imply a merger between CRM, marketing automation, website, email, and other customer-facing systems. However, you could have a centralized decision system to pick customer treatments while keeping separate delivery systems for each channel.
Q: Given your expertise in data strategies, what top three approaches should businesses adopt to get the most value from a CDP?
David: The most important data strategy is to be use-case driven: identify the most valuable use cases you can execute and assemble the data to make those possible. That’s very different from collecting every piece of data that you can because it might someday be useful.
A second important strategy is to concentrate on data quality and governance, ensuring the data you work with is as good as possible. This applies both to source systems and to data transformations within the CDP.
A third strategy is to deploy real-time capabilities whenever possible because these open up many new possibilities for better customer experience.
Q: How can organizations align their data strategies with a genuinely customer-first mindset?
David: Again, the primary focus needs to be use cases, which themselves should prioritize improving customer experience. Seeing things from a customer viewpoint is the best way to uncover opportunities to improve customer experience, which is the primary driver of business success. There can also be valuable opportunities unrelated to CX, such as reducing costs, but these are less strategic.
"Seeing things from a customer viewpoint is the best way to uncover opportunities to improve customer experience, which is the primary driver of business success."
Q: What criteria— scalability, integration, or pricing—are most critical when assessing a potential CDP?
David: Many factors are important. It would be a mistake to concentrate on one and exclude others. The Consultants Union requires that I say, “Every company is different. You have to look at each situation separately.” That said, our research shows that integration capabilities are generally the highest priority among successful buyers, followed by features and usability.
I will definitely say that price should be the last concern since there’s no value to a low-cost system that doesn’t fit your needs. People buy on price when they think all other things are equal about the candidate systems, or at least all candidates meet their needs equally well. But in fact that is rarely really true. Thinking the candidates are the same most often reflects an inadequate evaluation process.
Conclusion
As customer data continues to shape modern marketing, the role of CDPs has become more integral than ever. David’s insights highlight the critical importance of aligning CDP strategies with business needs, ensuring compliance, and preparing for AI-driven advancements.
Businesses looking to leverage CDPs effectively should focus on defining clear use cases, prioritizing data quality, and integrating real-time capabilities to enhance customer experiences. As AI and automation evolve, CDPs will remain a cornerstone of data-driven marketing, providing the foundation for personalized, seamless customer engagement.
Ultimately, organizations that adopt a strategic, customer-first approach to data management will be better positioned to navigate the future of marketing technology.
Follow David for more insights.
