Why composable CDPs are becoming the future of data management

Now that first-party data is the future of AdTech, publishers are gearing up to collect customer data. Heightened demand for customer data collection and processing brought CDPs into the spotlight.?

CDPs were an obscure concept a few years ago and struggled to differentiate themselves from DMPs, consent-management platforms, and other data solutions.?

Now, they are riding the Gartner hype cycle—partly because of the looming cookieless future and partly because the technology has come a long way.?

The shift from packaged to composable customer data platforms has been particularly appealing to publishers who no longer have to tweak their workflows to fit the features of the CDP.?

In this edition of the MadTech Digest, I explore what distinguishes composable CDPs from early platforms and why they are likely to be a cornerstone of publishers’ data stacks.?

If you are new to my newsletter and would like to get reminded when a new edition of the biweekly AdTech trend review or commentary comes out, follow MadTech Digest. I would also recommend reviewing the previous editions, such as my review of cookie alternatives.??

Back to composable CDPs.?

Why early CDPs did not take off

The concept of a CDP goes back to 2013, when David Raab , a marketing technology consultant and founder of CDP Institute , used it to describe Causata.?

The product, one of the pioneer CDPs, is no longer on the market, but its value proposition has shaped the industry's understanding of a customer data platform.?

What did Causata promise? In addition to standard database management, it touted data consolidation, aggregation, and the ability to build a single customer view through a robust integration ecosystem.?

The powerful vision of Causata’s team did not come to life because the tech needed to bring it to life had not yet matured.?

Specifically, early CDPs like Causata were missing:?

• The ability to support a lot of integrations. Data shows that a medium-sized team has between 5 and 10 platforms in their MarTech stack. Enterprises use over 120 marketing tools. In the 2010s, building API integrations for each one took so much time that it would have been unfeasible for CDP vendors. Restricted by the number of sources they could implement, companies did not get the single customer view they expected.?

• The need to send data directly to the platform limited the types of data CDPs could store. Thus, a common setup was a CDP for event data and a data warehouse for first-party audience data. The two were disconnected, and a single customer view remained a pipe dream.?

• Lack of market education. It’s only recently that CDPs are becoming an industry trend. In 2021, only half of MessageGears survey respondents correctly defined the purpose of a CDP—the ability to manage customer activities. There was (to an extent, still is) a lot of confusion about the features of a CDP and its place in the modern AdTech stack.?

Enter composable CDPs

CDPs may have had a rough decade before but have made a powerful comeback in recent years. What contributed to growing interest and demand??

The answer is composable CDP.

In a landscape where adaptability and customization in a MarTech stack are non-negotiable, the ability to move on from rigid closed-loop platforms to tailored solutions with flexible pricing turned the tide.?

Here’s how composable CDPs got the market talking:?

• Making the single customer view dream come true. A composable CDP infrastructure allows data to be imported from any source and accessed via SQL queries instead of custom-made APIs.?

• Technology-agnostic design: Companies can plug any data warehouse or data lake into the platform, eliminating vendor lock-in.?

• Flexible toolset. In off-the-shelf offerings, teams had to make do with the features they had, often missing out on use-case-specific capabilities. Composable CDPs allow teams to remove the features they do not need and add the tools they want (granular analytics, advanced identity resolution, or others). With modern-day CDPs, marketers need to rely less on data science teams to build new integrations and data pipelines.?

• Improved data governance. Composable CDPs eliminate the need for spreading data across databases, reduce duplication risks, and keep customer information within the organization’s security perimeter.?

What will come after composable CDPs??

Composable CDPs succeeded packaged CDPs, which, in turn, sidelined DMPs. It’s only logical that, in the near future, new technology will also substitute composable CDPs—and we already have an inkling of which one.?

The development of reverse ETL can increase the impact of data warehouses. Instead of focusing purely on data storage and being disconnected from the rest of the stack, data warehouses can use reverse ETL to share data with MarTech platforms. Census reverse ETL allows teams to build bilateral data flows between over 200 tools. Twilio also enriched its platform with reverse ETL.?

As data warehouses become more powerful, they encroach on CDP territory. Companies like Make are dropping CDPs in favor of a custom architecture.?

The good news for composable CDPs is that there’s still plenty of value to deliver. Building warehouse-native deployments into their platforms is one way to jump on the warehouse-native bandwagon. Building a robust suite of added features for conversion optimization, customer segmentation, and data activation is another strategy for staying relevant in a crowded marketplace.?

The bottom line is that whatever the future of CDP is, it will not be a monolith. Instead, we will have a wide range of mix-and-match platforms that allow publishers to handpick the features they pay for and maximize the value they get from their solutions.?

What are your thoughts on composable CDP's rise and the future in warehouse-native infrastructures?