95: Battle of the CDPs: Packaged vs. Composable, 10 Experts Weigh In

What’s up everyone, today we’re taking a deep dive into customer data and the stack that enables marketers to activate it. We’ll be introducing you to packaged customer data platforms and the more flexible options of composable customer data stacks and getting different perspectives on which option is best.

I’ve used both options at different companies and have had the pleasure of partnering with really smart data engineers and up and coming data tools and I’m excited to dive in.

Here’s today’s main takeaway: The debate between packaged and composable CDPs boils down to a trade-off between out-of-the-box functionality and tailored flexibility, with industry opinions divided on what offers greater long-term value. Key factors to consider are company needs and data team size. But if you do decide to explore the composable route, consider tools that focus on seamless integration and adaptability rather than those who claim to replace existing CDPs.

Jump to a Section 👇

The 8 Core Components of Packaged CDPs: What the Experts Say

Okay first things first, let’s get some definitions out of the way. Let’s start with the more common packaged CDPs.

A Customer Data Platform (CDP) is software that consolidates customer data from various sources and makes it accessible for other systems. The end goal is being able to personalize customer interactions at scale. 

I’ve become a big fan of Arpit Choudhury of Data Beats, he articulates the components of a packaged CDP better than anywhere I’ve seen in his post Composable CDP vs. Packaged CDP: An Unbiased Guide Explaining the Two Solutions In Detail

8 packaged CDP components:

  1. CDI (Customer Data Infrastructure): This is where you collect first party data directly from your customers, usually through your website and apps.
  2. ETL (Data Ingestion): Stands for Extract, Transform, Load. This is about pulling data from different tools you use and integrating it into your Data Warehouse (DWH).
  3. Data Storage/Warehousing: This is where the collected data resides. It’s a centralized repository.
  4. Identity Resolution: This is how you connect the dots between various interactions a customer has with your brand across platforms and devices.
  5. Audience Segmentation: Usually comes with a drag-and-drop user interface for easily sorting your audience into different buckets based on behavior, demographics, or other factors.
  6. Reverse ETL: This is about taking the data from your Data Warehouse and pushing it out to other tools you use.
  7. Data Quality: This refers to ensuring the data you collect and use is valid, accurate, consistent, up-to-date, and complete.
  8. Data Governance and Privacy Compliance: Ensures you’re in line with legal requirements, such as user consent for data collection or HIPAA compliance for healthcare data.

So in summary: Collect first party data and important data from other tools into a central database, id resolution, quality and compliance, finally having a segmentation engine and pushing that data to other tools.

I asked recent guests if they agreed with these 8 components 👇

Collection, Source of Truth and Segmentation

Boris Jabes is the Co-Founder & CEO at Census – a reverse ETL tool that allows marketers to activate customer data from their data warehouse.

When asked about his definition of a packaged CDP, Boris elaborated on the role these platforms have carved for themselves in marketing tech stacks. To him, packaged CDPs are specialized tools crafted for marketers, originally in B2C settings. Their primary utility boils down to three main functions: data collection, serving as a reliable data source specifically for the marketing team, and data segmentation for targeted actions.

The ability to gather data from various customer touchpoints, such as websites and apps, is crucial. These platforms act as the single source of truth for that data, ensuring that marketing teams can trust what they’re seeing. Finally, they provide the capability to dissect this data into meaningful segments that can be fed into other marketing tools, whether that’s advertising platforms or email marketing solutions.

Though Boris mentioned the term “DMP,” it’s essential to differentiate it from a CDP. Data Management Platforms (DMPs) have historically been tied to advertising and don’t provide that rich, long-term profile a CDP can offer. The latter offers a more holistic view, allowing businesses to target their audience not just based on advertising metrics but on a more comprehensive understanding of consumer behavior.

Hear directly from Boris below 👇

Key Takeaway: Packaged CDPs are functional units that collect, validate, and segment data for marketing utility. If you’re considering implementing an all-in-one CDP, look for these three core features: comprehensive data collection, a single source of truth for that data, and robust segmentation capabilities.

Adding Predictive Modeling to Packaged CDPs

Tamara Gruzbarg is the VP Customer Strategy at ActionIQ – an enterprise Customer Data Platform.

When asked about her stance on 8 components of a packaged CDP, Tamara generally concurred but added nuance to each element. Starting with data collection and ending with data activation, she emphasized the critical nature of these components. Tamara also advocated for the necessity of drag-and-drop UI for audience segmentation, which paves the way for data democratization and self-service.

Going beyond mere segmentation, Tamara revealed that her platform offers insights dashboards. These aren’t just Business Intelligence (BI) tools; they help marketers understand segment overlaps and key performance indicators, which further empower them to design more efficient campaigns. Her approach involves offering two types of audience segmentations: rule-driven and machine learning (ML) driven. The latter is a distinct component that allows clients to construct audiences based on predictive models, and it’s an option that has gained traction especially among mid-market businesses.

Tamara also touched upon a salient point regarding large enterprises. Even these giants can benefit from predictive tools when dealing with new data sets they hadn’t previously accessed. Collaboration with their in-house data science teams ensures the quality and reliability of this predictive modeling.

Hear directly from Tamara below 👇

Key Takeaway: A well-designed CDP should not just offer data collection and segmentation but also facilitate data activation and provide actionable insights. Whether you’re a large enterprise or a mid-sized business, the predictive modeling feature in some modern CDPs offers a fast track to gain valuable insights into your audience. Keep an eye out for these extended functionalities when evaluating a CDP for your business.

The Importance of Data Quality and Governance

Michael Katz is the CEO and co-founder at mParticle, the leading packaged Customer Data Platform.

When asked about his agreement with the often-cited eight components of a packaged Customer Data Platform (CDP), Michael did more than just nod in approval. He concurred that these elements are, at a minimum, the pillars of first-generation CDPs. Yet, he warned that very few platforms are strong across all these functionalities, giving his own platform as an exception for its comprehensiveness. According to Michael, a robust CDP is not just a collection of features but an integrated system where the entire value is greater than its individual parts.

Diving deeper into the conversation, Michael addressed a common shortfall in the CDP landscape—data quality and data governance. Many platforms, he noted, lack robust features in these areas. The result is an unstable foundation that undermines the value proposition of a CDP. In Michael’s words, the real magic happens when you can move from the data collection phase through to the data activation layer without compromising on quality and governance.

Michael also highlighted a nuanced point that often gets overlooked: the speed at which you can push data out into your application layer must be balanced with maintaining data quality and consumer privacy protection. It’s not just about how fast you can move; it’s about how fast you can move responsibly.

Hear directly from MK below 👇

Key Takeaway: When evaluating a CDP, don’t just look for a checklist of features. Look for an integrated system that’s strong in areas often neglected by others, such as data quality and governance. Speed is important, but not at the cost of quality and consumer privacy. Your CDP should offer more than just rapid data transfer; it should provide a stable, comprehensive platform for making that data actionable.

The Main Event: Harnessing and Activating Data

Tejas Manohar is the Co-founder and Co-CEO at Hightouch, another reverse ETL tool, that’s taken a bit more of a controversial stance.

When asked about the 8 components of a packaged CDP, Tejas broke ranks. He neither agreed nor disagreed with the elements but instead shifted the focus to the real question: Why do companies seek out a Customer Data Platform in the first place? According to Tejas, it’s primarily about harnessing and activating customer data to personalize experiences and drive better outcomes. Everything else, in his view, is ancillary.

In a field cluttered with feature lists and component breakdowns, Tejas urged companies to simplify. He distilled the CDP’s core functionality into three primary aspects. First, the platform must offer a mechanism for data collection. Second, it needs to provide some form of data transformation; think identity resolution and modeling. And third, it should facilitate data activation, typically through audience building and integrations.

For Tejas, the exhaustive lists of features and components often discussed in the martech space are merely a means to an end. Companies shouldn’t get lost in the weeds of features or components; instead, they should focus on what a CDP is fundamentally designed to achieve. Tejas argues that it’s not about ticking boxes on a feature list but about how these features contribute to the ultimate goal of using data effectively.

Hear directly from Tejas below 👇

Key Takeaway: Don’t get sidetracked by a long list of features or components when evaluating a CDP. Keep your eye on the main event: harnessing and activating data to improve customer experiences and business outcomes. Simplify your approach, and focus on the core functionalities that will help you reach your objectives.

So What’s the Hype Around Composable CDPs?

Thought of as the new kid on the block, composable CDPs promise a lot of different things compared to the packaged option. 

Composable CDPs take a modular approach to data management, built from separate, easily interchangeable parts. This design offers finer control over data processes and can be customized to fit particular business objectives. They provide a contrast to packaged solutions, balancing specialized benefits against workflow complexity.

Example tools/setup

  • CDI: Snowplow
  • ETL: Airbyte
  • DWH: BigQuery
  • Reverse ETL: Census
  • Data quality: dbt

But not everyone sees the composable route as an entirely new thing.

Drawing Parallels: Composable CDPs and the Lessons from Headless Commerce

David Chan is Managing Director at Deloitte Digital and leads their CDP practice.

When asked about the buzz surrounding composable Customer Data Platforms (CDPs), David turned the spotlight on a parallel from his own background—headless commerce. Originating around 2013-2015, headless commerce was a game-changing moment that separated web content management from the commerce tools themselves. In this setup, the content management system functioned as the front end, while the commerce tools handled the heavy-duty logic like checkouts and product details. David observed that this shift towards modularity in the commerce space was an early sign of how composability could transform industries.

David then dissected the current state of composable CDPs, comparing it to the early days of headless commerce. The crux of the issue, he said, lies in integration. While today’s CDP landscape is flush with features and capabilities, it’s noticeably lacking a unified framework for how these components should interact. This fragmentation echoes the initial phases of headless commerce, where disjointed systems eventually gave way to more standardized, interoperable solutions.

What sets the CDP space apart right now, according to David, is the absence of those well-defined standards and partnerships that can guide the development of composable architectures. The commerce space underwent a similar period of “mashing and banging,” where different features and tools were reluctant to work in concert. Eventually, standards emerged that dictated how these composable elements should fit together. This level of structure, David argues, is still conspicuously absent in the world of CDPs.

Hear directly from David below 👇

Key takeaway: Composable CDPs are still in their formative stages. But given the trajectory witnessed in headless commerce, it’s only a matter of time before these platforms evolve to include more standardized, collaborative frameworks. That’s what will take them from being a collection of features to a cohesive system, just like headless commerce did years ago.

Tap Into Existing Data in Your Warehouse

When asked about the shift toward composable CDPs, The Co-CEO of Hightouch explained that while the demand for CDPs is high, the satisfaction derived from most available solutions leaves something to be desired. Tejas cited a Gartner report indicating that a mere 60% of organizations find their CDPs valuable. The issue isn’t with the CDP concept, but rather with its traditional execution of making a copy of your data—hence the rising interest in composable CDPs.

Tejas contends that composable CDPs offer a much-needed alternative. These platforms are designed to tap into the extensive data already stored in an organization’s data warehouses. This approach integrates existing data pools, breaking down data silos, and making it accessible to marketing teams. The result is a more practical and efficient way to activate personalized customer journeys.

The push toward composable CDPs, then, isn’t just a passing fad. It’s a meaningful evolution aimed at resolving real-world dissatisfaction with older CDP models. By enabling marketers to seamlessly leverage existing organizational data, composable CDPs stand to make the concept of a CDP not just aspirational but genuinely functional.

Hear directly from Tejas below 👇

Key takeaway: The movement toward composable CDPs is rooted in the need for a different data architecture and utilization of existing data. While traditional CDPs sometimes fall short of delivering on their promise, composable CDPs aim to make existing organizational data accessible and actionable for marketers.

The Need to Adapt to Complex Customer Journeys and Regulatory Demands

When asked about the factors motivating the industry’s move towards composable Customer Data Platforms (CDPs), the Co-founder and CEO of Census explained that it wasn’t merely a matter of opposing the traditional CDPs. Instead, the focus was on first principles, aiming to provide marketers with more trustworthy data. Boris emphasized that existing data storage solutions, like data warehouses from Google, Snowflake, Amazon, or DataBricks, already hold extensive and infinitely flexible data sets. The question then becomes, why duplicate these resources?

Composability, Boris shared, isn’t about disassembling systems but about creating components that can seamlessly work together. This approach allows businesses to customize parts of the system without disrupting its overall functionality. Traditional CDPs tend to fall short because they can’t offer the level of flexibility modern businesses require, particularly as customer journeys become more complex and multi-faceted.

Boris also discussed the increasing complexity in customer journeys, pointing out that marketing has evolved significantly from the times when placing a pixel on a website would solve most tracking issues. Today, especially in the B2B sector, customer relationships and touchpoints are more varied and complicated than ever before.

Lastly, Boris touched on emerging regulatory demands. Marketers now have to navigate complex privacy requirements. Whether it’s the EU or California, companies are expected to be more transparent about data collection, storage, and usage. This shift makes first-party data and its proper governance crucial, adding another layer of complexity to an already intricate landscape.

Hear directly from Boris below 👇

Key takeaway: The shift towards composable CDPs isn’t just a reactionary move against traditional platforms; it’s an evolution driven by a need for more reliable data, increased flexibility, and the capability to adapt to complex customer journeys and regulatory demands. By focusing on composability, companies can harness their existing data infrastructure to build more agile, adaptable systems.

Debating the Merits of Composable Versus Packaged CDPs

Okay so we’ve covered the components and the definition of a packaged CDP and why there’s a need for some companies to explore the more flexible route of a composable stack. 

Let’s hear from various different industry pros about where they side when it comes to the packaged vs composable CDP battle 👇

Choosing the Right Customer Data Platform: Flexibility vs. Cost in the CDP Debate

Wyatt Bales is Chief Customer Officer at Bluprintx, a global Growth-as-a-Service consultancy who provide Martech, Salestech, and Work Management solutions.

When asked about his stance on the debate between packaged and composable Customer Data Platforms (CDPs), Wyatt shed light on some crucial considerations. He noted that for some companies, the ongoing licensing costs of Segment have become a long-term burden. Wyatt referenced a data integrity customer in Belgium, as an example of a company benefiting from a different approach. Instead of operating on a traditional CDP, they use a data warehouse like Snowflake to gauge the quality of their data.

Wyatt emphasizes that modern data warehouses have evolved to offer a wide array of tools. These tools, which sit atop the warehouse, serve as insightful indicators of what kind of data you’re dealing with. Whether it’s about understanding data cleanliness or complexity, the warehouse can act as a hub for diverse data operations. This kind of flexibility makes warehouses an increasingly attractive option for companies looking beyond traditional CDPs.

The discussion then veered into the realm of API calls for tasks such as email delivery and campaign execution. Wyatt is convinced that, particularly for the enterprise space, the future lies in leveraging data warehouses for these outbound tasks. The inherent adaptability of warehouses allows for easier integration of various functionalities, offering a nuanced, practical approach to handling customer data.

Hear directly from Wyatt below 👇

Key takeaway: It’s not about choosing one type of CDP over another but understanding your specific needs and options. Companies may find that the flexibility and scalability of modern data warehouses make them a suitable, if not superior, alternative to traditional CDPs.

📫 Never miss an episode or key takeaway 💡

By subscribing to our newsletter we’ll only send you an email when we drop a new episode, usually on Tuesday mornings ☕️ and we’ll give you a summary and key takeaways.

Success! You're on the list.

Why Data and Messaging Integration Matters in the Packaged vs Composable Debate

Pini Yakuel is the CEO of Optimove, a platform that combines a Customer Data Platform (CDP), a journey orchestration tool and an AI engine.

When asked about the ongoing debate between packaged and composable CDP and martech solutions, Pini emphasized the importance of contextualizing each company’s unique needs. He argued that for many businesses, the true value of data lies in its proximity to messaging channels. In this setup, data isn’t just a dormant entity waiting for analytics; it actively informs real-time decisions to improve customer interactions.

Pini pointed out a common pitfall: the fragmentation of data and channels. While some tools may excel at data management, they often export that data to another tool responsible for messaging. This can create a disconnection between data analytics and actionable insights. The exported data is fed into a system that remains, at its core, rule-based rather than data-driven.

Diving deeper into the importance of integrating data with decision-making, Pini indicated that when data and channels share the same platform, they enable an “AI feedback loop.” This is not just about smarter segmentation; it’s about making the entire system inherently smarter. An integrated platform can be adaptive, not just reactive. Such a setup cannot be easily replicated by stringing together APIs from different systems because the latter approach doesn’t change the fundamental nature of those systems—they remain rule-based.

Closing out his argument, Pini revealed the mindset driving his company’s approach to solving this issue. To truly unlock the power of data, they maintain an “obsession” with solving this particular problem. Their unwavering focus enables them to slowly piece together a more comprehensive and optimized solution where data and channels coexist in a virtuous cycle.

Hear directly from Pini below 👇

Key takeaway: The debate between packaged and composable CDP and martech tools isn’t about one being universally better than the other. It’s about understanding that the real power comes from aligning your choice with your specific needs and goals, especially when it comes to integrating data and messaging channels for actionable insights.

Choosing Flexibility and Innovation: The Case for Composable over Packaged

Arun Thulasidharan is the CEO & Co-founder at Castled.io – A warehouse-native customer engagement platform that sits directly on top of cloud data warehouses.

To him, the core difference between composable and packaged CDPs resembles the contrast between open source and closed source systems. A composable CDP, built atop a data warehouse, bestows the flexibility to innovate. If you find something lacking, you’re not confined; you can add more tables or transformations to the system.

Arun emphasizes that this flexibility is not just theoretical; it’s practically beneficial. He brings into play real-world examples, citing tools that perform identity resolution on top of a data warehouse. These tools employ fuzzy logic, rather than deterministic methods, to identify that two rows of data might actually be related. In doing so, they enable a new kind of innovation—one that can only occur in an open system, directly on the data warehouse.

In contrast, packaged CDPs often restrict this level of flexibility. They operate in a closed system, limiting your ability to introduce new functionalities or plug in external tools. To Arun, this lack of adaptability can stifle the innovations that are currently shaping the martech industry.

Yet, Arun acknowledges that the discussion isn’t black and white. There are compelling arguments for both sides, but his preference leans toward the composable model for its adaptability and the freedom it offers for innovation.

Hear directly from Arun below 👇

Key takeaway: Flexibility is currency in today’s martech landscape. Opting for a composable CDP over a packaged one can provide you the elbow room to innovate and adapt, positioning you at the forefront of industry advancements.

Cloud Data Warehouses, Data Strategy, and the Real Value of CDPs

Once again, let’s get thoughts from Michael Katz, the CEO of mParticle (packaged CDP) about how he genuinely feels about the packaged vs composable CDP debate.

MK asserted that the dialogue around it is often reduced to noise—distracting from the core issue. To him, the evolution of CDPs is not a luxury but a necessity, paralleling the demands of any growing business. He cut through the chatter to highlight the critical role of Cloud Data Warehouses, noting they serve as an organization’s single source of truth, at least in theory.

However, MK acknowledged that simply setting up a data warehouse doesn’t solve all problems. He emphasized the critical need for a robust data strategy and mechanisms to ensure data quality and integrity. The challenges don’t stop at data collection; they extend to navigating an ever-changing landscape of privacy regulations. MK clarified that the value provided by legacy CDP vendors like mParticle is not merely in data storage but in the movement and activation of data.

MK also argued that the real evolution in CDPs is away from basic segmentation tools toward more nuanced ‘journey tools.’ These not only collect data but offer a greater understanding of that data—providing context and insights. He shared that his focus over the past year and a half has been to move beyond just verifying the data’s truth to finding its meaning. Whether it’s looking back to understand what happened or looking forward to predict future outcomes, the goal is an ‘infinitely optimizing loop.’

Not mincing words, MK criticized the strategy of companies offering reverse ETL solutions. He labeled their approach as “garbage in, garbage out,” cautioning that a quicker path doesn’t equate to better results if you have garbage data in the first place. He also tackled what he perceives as distracting tactics—myths about zero data copy, unfounded security concerns, and misleading narratives on deployment times. MK pointed out that initial value and sustained value are not the same; what is easy to initiate is often difficult to maintain in the long run.

Hear directly from MK below 👇

Key takeaway: Discarding the noise is the first step to understanding the real value of CDPs. It’s not just about having a data strategy; it’s about continuously refining it to move from data storage to data activation and insights. MK warns against the allure of quick fixes and emphasizes that true value in the data space is a long game, demanding a robust strategy and the right tools.

Understanding the Distinct Roles of CDPs and Reverse ETLs in Marketing Strategy

Pratik Desai, is the Founder and CEO of 1to1, a personalization agency that works with enterprise clients and has recently released a product called Ragana, a composable search and sort personalization engine built on top of your eComm platform. 

When asked about the tension between packaged and composable Customer Data Platforms (CDPs), Pratik identifies a core issue: the debate often stems from a misunderstanding of what Reverse ETL tools are actually doing. He explains that marketers are sometimes sold on Reverse ETLs as if they’re a one-for-one substitute for CDPs. That’s misleading. Reverse ETLs and CDPs are solving different problems. CDPs, for instance, excel in identity resolution, a feature Reverse ETLs don’t offer.

Pratik digs deeper into the structural gaps that led to the rise of CDPs. Historically, marketing teams were often sidelined when it came to data strategy. CDPs emerged as a tool to give marketers a “seat at the data table.” Reverse ETL tools have value, but they won’t inherently solve this organizational disconnect. Buying a new tool won’t suddenly align your marketing and data teams if those teams weren’t aligned in the first place.

Switching focus to enterprise-level challenges, Pratik highlights the importance of operational excellence and data structure. The adoption of a reverse ETL tool won’t automatically resolve operational inefficiencies or integrate marketing into broader data strategy. It’s not a silver bullet for organizational issues.

Pratik ends by urging businesses to examine their unique problems before leaping into any tech solutions, whether it’s a CDP or a Reverse ETL. Some organizations, particularly SMBs where marketing already has data influence, can extract enormous value from reverse ETL tools. However, Pratik warns that we’re a long way from a one-size-fits-all solution, especially for enterprise-level customers.

Hear directly from Pratik below 👇

Key takeaway: Understand your organization’s specific challenges before diving into CDPs or Reverse ETLs. These tools are not interchangeable; they solve distinct problems. Align your teams and clarify your data strategy first—only then can you effectively leverage these technologies.

Can Reverse ETL Really Replace Packaged CDPs?

So let’s talk about the confusion in the market. Can reverse ETL actually replace a packaged CDP?

Most of the confusion stems from one Reverse ETL vendor in particular: Hightouch. They’ve written plenty of controversial articles claiming that the CDP is dead and that they can replace it. I sat down with Tejas, the Co-founder and Co-CEO to get to the bottom of why they think they can replace the packaged CDP.

Is the CDP Really Dead?

Tejas mentioned that large Hightouch customers like Blizzard Activision and Warner already refer to their platform as a CDP (at elast internally). But what makes Tejas’ perspective intriguing is that their product doesn’t fit the typical mold of a CDP.

The core of Tejas’ viewpoint rests on the activation of marketing data. He emphasizes that the ultimate differentiator in this space isn’t just the collection of data, but how effectively a company can activate and personalize that data. Tejas hints that traditional CDPs often fall short in this area. While they collect mountains of data, they lack in providing actionable insights and seamless data activation for marketing teams.

Tejas went on to address a bold prediction made in their company’s blog post, stating “CDPs are dead.” He argues that the CDPs of the future will either adapt to the flexible, data activation-centric model their company has pioneered or risk becoming obsolete. In Tejas’ eyes, they are shaping the future of the CDP landscape by focusing on what matters the most—enabling companies to own their data, offering infinite flexibility, and allowing data activation across all channels.

Hear directly from Tejas below 👇

So some large customers do refer to Hightouch as their internal CDP. Is that enough to be able to claim that the CDP is dead and that they can replace it? 

I asked Michal Katz for his take on Tejas’ argument.

Challenging the DIY Approach to CDPs

When asked about Tejas’ claim that some customers refer to their platform as an internal CDP, MK offered a nuanced take. He argues that this viewpoint represents a narrow segment of the market. For Michael, the fragmented DIY (Do It Yourself) approach, often favored by data engineers, falls short in delivering business value, especially for enterprises. MK warns that the “day of reckoning is coming,” as sloppy habits have been formed, particularly during the pandemic. According to MK, these habits often stem from data engineers operating without proper business requirements, resulting in suboptimal digital marketing campaigns.

MK points out a significant shift that has occurred over the past 9-12 months—marketers are reclaiming power from data engineers. mParticle is built with marketers in mind, focusing on low-code or no-code data activation. MK notes the importance of usability in delivering value, contrasting their approach with some of the more complex CDPs. Hes emphasize that their platform allows for easy data contextualization and activation, all through a point-and-click user interface.

Beyond usability, MK makes a case for integrated platforms, pushing against the notion of using different components for an end-to-end CDP. He highlights the challenges of troubleshooting across multiple systems, especially when things go awry. According to MK, managing across different platforms introduces unnecessary complexities and slows down the ability to deliver business value.

MK concludes by stressing that while there may be many paths to value creation, the quickest isn’t always the most straightforward. In his view, there’s considerable value in using an integrated platform where seamless workflows are a given and customization is an option but not a necessity.

Hear directly from MK below 👇

Not All Reverse ETL Tools Aim to Replace the CDP

Not all Reverse ETL platforms have taken a loud controversial approach to marketing though. I asked Boris Jabes, the CEO and Co-founder of Census if his platform replaces a CDP, he emphatically said no.

He explained that many of his customers use his product, Census, alongside a traditional CDP. Far from making these platforms obsolete, Boris’ goal is to give users access to trustworthy data across multiple locations. His focus is on composability—a philosophy that emphasizes building tools that seamlessly integrate with existing systems. Rather than adding another data silo, Census aims to utilize a business’s existing data infrastructure.

Boris took issue with the cutthroat language often seen in brand marketing—this idea that one tool “kills” or “destroys” another. According to him, composability benefits everyone; it’s the unsung virtue that ensures different tools can work together without causing chaos. This isn’t just a win for the marketing team. Sales, finance, privacy, and compliance—all can leverage the same cohesive data structure.

Boris also noted that composability isn’t just about making it easier for marketers. It’s a guiding principle in software development, often discussed even at the level of programming languages. Whether you’re a marketer or part of a data team, he advocates for tools that don’t just serve their isolated purpose but can also integrate effectively with other components of a business’s tech stack.

On the topic of identity resolution, Boris argued that if it’s happening in your CDP, that shouldn’t be exclusive to the marketing team. Census aims to democratize access to this crucial data, ensuring it benefits the entire organization. This is not about Census trying to replace CDPs; it’s about working harmoniously with them to provide a well-rounded, integrated solution.

Hear directly from Boris below 👇

Key takeaway: The question isn’t whether Reverse ETL can replace a CDP, but how these tools can coexist and complement each other. Composability is the bridge that allows for this harmonious relationship, making the data landscape more functional and less complicated for everyone involved.

Why the Idea of Reverse ETL Replacing CDPs is Misleading

David Chan, the Managing Director at Deloitte Digital who leads their CDP practice is really close to the composable vs packaged debate. Despite wanting to move on from the debate, I asked him about the perspectives of Census and Hightouch.

David pointed out his understanding of Boris’ perspective, stating that Boris wasn’t aiming to dethrone CDPs but rather to create robust tech solutions for data teams. In contrast, he questioned the sincerity behind Tejas’ claim that a Reverse ETL tool could take the place of a CDP.

David pulled apart the anatomy of a CDP to make his point. He sees Reverse ETL as just one piece of the larger CDP puzzle. Arguing that no single tool should claim the full functionality of a CDP, David raised some important questions: Is a Reverse ETL tool responsible for real-time and batch data collection? Does it handle transformations and all inbound ETL? Is it responsible for identity resolution? His answer was an emphatic no. He suggested that what Tejas and Hightouch are offering is valuable but shouldn’t be inflated into something it’s not.

He went on to say that what Hightouch and similar platforms can do is to integrate seamlessly into native enterprise data warehouses. This enables frontend users to query data, build audiences, and connect to various downstream systems in martech and ad tech. However, this isn’t the same as serving as a comprehensive CDP solution.

David seemed to imply that the idea of Reverse ETL replacing a CDP might just be a marketing gimmick to generate buzz. He did acknowledge that such tools offer a convenient plug-and-play into enterprise data structures but felt it was misleading to label them as CDP substitutes.

Hear directly from David below 👇

Key Takeaway: The dialogue shouldn’t focus on whether Reverse ETL can replace a CDP, but rather how it functions as a component within the broader data ecosystem. Recognizing the limitations and specific utilities of each tool will lead to a more effective and truthful martech strategy.

The Irony of Reverse ETL Tools Possibly Becoming the CDPs They Oppose

Since we last spoke to Tejas, Hightouch added ID resolution and event collection features. While they are built on your warehouse vs copying your data, many are arguing that this is making them start to look a lot like the packaged CDP they initially claimed to be dead.

Aliaksandra Lamachenka, a data and martech consultant is a big fan of the composable architecture for some companies depending on their stage and their team but she shared some of her thoughts on the evolution of CDPs.

Aliaksandra candidly dismissed the notion of a one-size-fits-all solution. She believes that as businesses grow, their needs for functionalities will also evolve. Vendors then face a critical decision: either evolve alongside their customers by adding functionalities, or stay true to their core offerings. According to Aliaksandra, this fork in the road could have significant industry implications.

Interestingly, she noted that adhering to core values and functionality could actually be beneficial for the industry. Such a stance supports market democratization by serving a broader range of customers at different stages of growth. However, she also points out that vendors will inevitably reach a juncture where they must decide whether to keep adding layers to their platforms to meet customer demands, or to specialize and remain focused on their core offerings.

The discussion on packaged vs composable CDPs is happening in a vacuum, Aliaksandra feels. While the industry debates the merits of one over the other, companies are struggling with more immediate and foundational issues like data quality, data lineage, and system discrepancies. For her, these problems underscore the necessity of having a strong data infrastructure in place before even considering which type of CDP to adopt.

Aliaksandra highlights a common pitfall: the industry’s fascination with adding new tools without considering the state of the existing data. She argues for the importance of first having a “clean” data layer to build upon. Without it, no CDP—packaged or composable—can be fully leveraged. In essence, she champions the idea of data hubs that enforce ownership and documentation by design as a foundational step.

Hear directly from Aliaksandra below 👇

Key takeaway: Before diving into the packaged vs composable CDP debate, focus on the basics. Ensure your data is in good shape and avoid adding new tools to a chaotic environment. Once your data is well-managed and reliable, you’ll be in a position to make more informed choices about which type of CDP best suits your needs.

Episode Recap

So there you have it folks, Hightouch believes they can replace a packaged CDP because a few of their customers refer to them as their internal CDP. But Michael Katz thinks that represents a very narrow segment of the market. He argues that a fragmented DIY approach doesn’t optimize business value and isn’t practical for most enterprises. 

On the other hand, Boris and Census have taken a less controversial approach to product marketing and opted for more honest messaging. They don’t claim to replace CDPs, in fact they’re happy to work alongside them. The core idea behind Census is creating tools that integrate well with existing systems without adding complexity. Census seeks to distribute trustworthy data across different departments, leveraging existing data infrastructure rather than creating another silo. 

David seems to side with Census on this part of the debate. He doesn’t think reverse ETL tools can replace CDPs entirely. His view is that reverse ETL tools are just one component of a complete CDP system and claiming that a reverse ETL tool could serve as a CDP would be misleading. 

Finally, Aliaksandra argues that vendors offering composable, lightweight solutions are making data more accessible. However, these vendors face a strategic choice: either expand their feature sets to keep customers engaged, risking the credibility of their original bold statements like “the CDP is dead,” or focus on perfecting their core offerings without overhyping their capabilities.

You heard it here first folks: The debate between packaged and composable CDPs boils down to a trade-off between out-of-the-box functionality and tailored flexibility, with industry opinions divided on what offers greater long-term value. But if you do decide to explore the composable route, consider tools that focus on seamless integration and adaptability rather than those who claim to replace existing CDPs.

Listen to the full episode 🎧


Intro music by Wowa via Unminus
Cover art created with Midjourney

See all episodes

📫 Never miss an episode or key takeaway 💡

By subscribing to our newsletter we’ll only send you an email when we drop a new episode, usually on Tuesday mornings ☕️ and we’ll give you a summary and key takeaways.

Success! You're on the list.

Future-proofing the humans behind the tech

Leave a Reply