It’s been a little over three years since Raab Associates introduced the label “Customer Data Platform” to describe an emerging class of marketer-controlled systems that use persistent, cross-channel customer data to support external marketing execution. Adoption of the term was low at first but has now accelerated sharply. A recent Google search found at least 18 vendors positioning themselves as CDPs compared with just a handful a year ago.
Beyond broader use of the term itself, here are some other trends from the past year:
Greater diversity in CDP outputs.
All products calling themselves CDPs assemble customer behaviors from multiple source systems, create a unified profile, and share the results with other marketing systems. But they differ substantially in how this sharing happens. Some allow direct access to the profile database via queries or APIs. Some push the profiles to an external database. Some attach segment codes or predictive model scores to lists that are exported to other systems. Many offer more than one of these options, the better to meet a broad range of marketer needs.
Movement towards higher-value output.
A “pure” CDP would only assemble customer profiles and passively make them available. But an increasing number of systems also run campaigns to create segments or even select messages for specific individuals. Outputs might be audience lists for display advertising, address lists for email delivery, or HTML messages to display on a Web page. Selection of lists or messages gives marketers an immediate return on their CDP investment, something that building an accessible database by itself does not.
Continued separation of CDPs from delivery systems.
Even when CDPs select marketing messages, they still stop well short of replacing message delivery systems such as ad servers, email engines, or Web content management. Those are complex operational systems that most marketing departments are reluctant to replace and the CDP vendors see no reason to reinvent.
Growing recognition of the need for a separate CDP.
Marketers are acutely aware that they need a unified customer database to support coordinated interactions across channels. Many have discovered that the major “marketing clouds” don’t deliver advanced data integration or flexible data storage of a CDP. Nor do the cloud systems necessarily give external tools easy, unlimited access to their data. These marketers have increasingly opted for a modular, best-of-breed approach that lets them use whatever tools they want with shared customer data in a CDP as the common connector. If the CDP can coordinate and select marketing messages, so much the better – but the real goal for these marketers is simply to avoid being locked into any one vendor’s collection of products.
Convergence between CDP and DMP (data management platform).
The distinction has always been clear: CDPs work with both anonymous and known individuals, storing “personally identifiable information” such as names, postal addresses, email addresses, and phone numbers, while DMPs work with almost exclusively with anonymous entities such as cookies, devices, and IP addresses. Indeed, anonymity is essential to the DMP’s role as a way to exchange information about audiences without violating personal privacy. What’s changed is that CDPs are integrating more often with advertising systems, and thus storing more DMP-type information such as cookie IDs with audience tags. Some DMPs are also storing personal identifiers, although these are carefully isolated from situations where anonymity is still important. But just adding personal identifiers doesn’t give a DMP the advanced identity matching and flexible data storage built into CDPs. So it will be hard for most DMPs to match full CDP functionality.
Extension to content data.
CDPs will not become content management systems. But more precise personalization means marketers must create vastly larger quantities of content and must capture more information about each item. Managing data about content is therefore an increasing headache that in some ways resembles the challenge of managing data about customers. The relationship becomes closer when marketers need to understand how each piece of content relates to specific individuals. Some CDP vendors have begun to meet these needs by adding content classification and matching features in their systems. More can be expected to follow.
What about the future? It’s safe to predict that CDP adoption will accelerate as marketers understand their data needs more clearly and become more aware that CDPs offer a solution. It’s also reasonable to expect that the trends listed above will continue. Things get murky once you look beyond current trends but here are some new things you might see in the next year or two:
More direct data access.
Once they’ve built their CDP, marketers will be eager to find new uses for it. This means they’ll want to give more systems direct access to the raw CDP data and to lightly processed versions such as aggregates, indexes, and derived data values. Direct access will increasingly be part of buyers’ requirement lists and the CDP vendors will respond by making their data more available. (Or, more precisely, those whose technology permits this will do so. Not every CDP can support direct access.)
Closer integration with external sources.
The CDP model is based primarily importing data from external systems and merging it into one big database. That will remain the main approach because marketers need the control it provides over how data is manipulated and stored. But there is an increasing number of external databases that cannot be imported for practical reasons, such as size or volatility. Fast-changing, location-specific data such as weather conditions is good example: you only need a tiny fraction of what’s available so it doesn’t make sense to try to capture it all. CDPs will get better of connecting with such sources on an as-needed basis.
Easier to use.
CDPs are by definition designed to be used by marketers, not IT specialists. But the current products still require considerable technical skills to set up and maintain. As interest in CDPs spreads beyond early adopters, the vendors will strive to reduce the time and skills needed to set up, maintain, and operate their products. Beyond better user interfaces, this could include prebuilt integrations with source and delivery systems. It might even extend to artificial intelligence-driven features that automatically incorporate new data elements with minimal human involvement.
More vertical systems.
There are already vendors using the CDP label who specialize in mobile marketing and programmatic ad analysis. As the market expands, other specialists will likely emerge with products for specific industries, small or mid-size companies, or service agencies. Each set of marketers has somewhat different needs, so a CDP tailored to their segment could be easier to deploy and use. Dominating a chosen niche becomes increasingly important as vendors face more competition.
Broader and less precise use of the CDP label.
As the term CDP becomes more popular, it will be adopted by vendors who don’t really meet the definition. Such is the price of popularity. Marketers will need to look closely at what each “CDP” vendor does, to make sure it matches their needs and expectations.
There have been surprisingly few acquisitions of CDPs to date. This will almost surely change as “marketing cloud” vendors recognize that the tools they acquired to build their suites were never designed to do what customers want: build a unified, multi-source customer database accessible to all external systems. Since this is what a CDP delivers, buying one is the obvious way for vendors to fill that gap. So don’t be surprised to hear that a big marketing software vendor has purchased a CDP – or that other vendors follow that first buyer with a chain of me-too acquisitions as they strain to keep pace.