Believe It or Not, the CDP Industry Is About to Get Less Confusing
October 15, 2024Today’s news brought an intriguingly complementary pair of announcements: one from long-established CDP vendor Lytics, that it has decided to “transform” the CDP category by defining an “Experience CDP” (xCDP) that includes personalization, and one from personalized messaging vendor MessageGears, that it is adding a full set of CDP capabilities. The scopes of the two expanded products are not identical: Lytics has a unique integration with WordPress and Drupal, while MessageGears actually sends email, SMS, and other types of messages which Lytics does not. But the underlying vision and language used by both vendors are nearly identical: in both cases, they see the need for a single system that assembles and activates customer data.
There’s nothing particularly new about this type of product expansion. Many CDP vendors that started out as tools to build shareable profiles later added journey orchestration and message selection capabilities, and many vendors that started as personalization or delivery tools have added the data assembly features needed to be a CDP. In fact, such broad-scope tools have long dominated the CDP industry: the CDP Institute’s latest Industry Update report shows that what we classify as “campaign” and “delivery” CDPs account for two-thirds of all vendors and three-quarters of industry employment, figures that have barely budged in the past five years. Quite simply, the market has spoken and it prefers integrated solutions.
What today’s announcements illustrate, then, is the continued convergence of most CDP vendors on a full set of functions ranging from data collection to personalization. Lytics argues that labeling this an “Experience CDP” (and, presumably. setting it as a new standard that all CDPs should meet) would somehow reduce confusion about CDP definitions. I personally don’t find this convincing, since Lytics’ definition of “Experience CDP” is quite similar to what the CDP Institute calls a “campaign CDP” (which, despite the misleading name, includes real time interactions as well as outbound marketing campaigns). One apparent difference is that Lytics includes tight integration with content management systems among its requirements, although you might argue that is implied in any system that manages inbound and outbound marketing campaigns.
In any event, given that the majority of CDPs already fall into the “campaign” and “delivery” categories, it seems that promoting an “Experience CDP” is refighting a battle that’s already been won. As even the most casual industry observer knows, the real debate within the industry recently has been between traditional CDP vendors, who provide complete, integrated systems to build customer profiles, and “composable CDP” vendors, who offer individual components for particular CDP functions.
But that’s another battle which is already over. While most “composable CDP” vendors started out offering a single function (reverse ETL), composable leaders Hightouch, Census, Rudderstack, and GrowthLoop now all provide a more-or-less complete set of the data capture, transformation, and unification functions needed to build customer profiles. When they sold just one component, those vendors argued that buyers really prefer to assemble their own solutions from independently-sourced components. But now that they have multiple components, they’ve discovered the benefits of buying pre-integrated components from a single vendor. (So far, the main composable vendors have not extended their systems to include substantial personalization capabilities, but you can bet they’re working on it.)
At the same time, many of the traditional CDP vendors have split their products into modules that can be purchased as separate components. So, while the composable vendors have been converging their functionality with traditional CDPs, the traditional CDPs have been converging their architecture to support a component-based approach.
The same dynamic applies to “warehouse native” architectures, the position that most composable CDP vendors have adopted as they move away from touting single-function components. While this has been a major selling point for the composable CDPs, many of the traditional CDPs have also adopted their products to read directly from cloud data stores such as Snowflake, Databricks, Google BigQuery and Amazon Redshift. Some can use the cloud data stores as their primary CDP database; others can only read supplemental data from the cloud systems. Either way, the distinction between traditional and composable architectures is increasingly unimportant.
All this convergence might seem to imply an increasing sameness among CDP vendors, but that’s not the case. The real divide within the industry is between vendors who provide IT teams with the tools to build customer profiles, and the vendors who provide marketing teams with a complete set of personalization capabilities (including customer profiles). This is primarily a functional distinction, although selling to different groups also implies different marketing, sales, pricing, and product strategies. This means we can expect two fairly distinct sets of vendors to emerge, each targeting one market or the other.
IT buyers might prefer separate components over integrated packages, and will surely prefer warehouse-native solutions over separate CDP databases. This makes selling profile-building tools to IT teams a natural fit for composable CDP vendors. But selling to IT is challenging because there’s stiff competition from well-established data tool vendors like Informatica and Precisely and from the newly componentized and warehouse-friendly traditional CDPs. So don’t be surprised if at least some of the composable CDP vendors continue to expand their functional scope in an attempt to support marketers instead. Those vendors are indeed likely to end up looking a lot like what Lytics has called an Experience CDP.
In the end, then, the CDP industry will get less confusing. The current architectural debate over integrated vs component- or warehouse-based architectures will fade to irrelevance, and CDP vendors will fall into only two categories: profile-builders and personalization platforms. This won’t necessarily make buying a CDP any easier, since differences among individual products will still be important in finding the right fit for each situation. But we can at least hope for clarity over what kind of CDP a company is looking for and less noise about which kind of CDP is best.