If marketers each had a single comprehensive profile that followed them around to every conference and client meeting they attended over the last six months, a common data point in these profiles, across the marketing tech industry, would likely be the much-discussed CDP, or Customer Data Platform. It’s on everybody’s mind, even though there is little agreement on what a CDP actually is. Sure, we understand the need to recognize individual customers and to quickly make sense of where they are on their journeys, all toward figuring out how to help them move forward to the next step in the process. This grasp of customer data is essential, and marketers require it to be accessible throughout their stack in an orderly, fluid way. The confusion comes when any number of related solutions profess to be CDPs out of fear that they will be passed over if they don’t use the proper buzzword.
We’ve seen in recent weeks how major vendors in the space are speaking to the needs while struggling with and revising the language. Kim Davis reported from this week’s Adobe Summit on how the new Adobe Experience Platform is billed as a “real-time CDP,” and he drew comparisons to what he’d previously heard from Oracle. He surmises: “At best, the big players in the customer experience space are facing up to the fact that seamless journeys can’t be delivered unless there is one, and just one, up-to-date customer profile. At worst, it sometimes seems they are trying to trump the entrants in the Customer Data Platform game.”
On cue, Winterberry Group released new research on CDP capabilities and semantics. One impression from the more than 100 companies they spoke with is that marketers know what a CDP should be. The confusion comes from not knowing when it’s not one.
Winterberry Group managing director Michael Harrison told me he’d never seen a topic like this that lacks so much clarity. Solutions that get roped in with CDPs “all came from different origins, from campaign management teams, personalization engines, orchestration hubs, platforms that are historically trying to save big data problems, and solving complex data issues and allowing you to activate that data to any application,” he stated.
According to Harrison, many marketers have it at the top of their list in 2019 to get a CDP. “But if you don’t have a definition,” he said, “software isn’t going to solve the problem.”
There’s nothing controversial in the narrowed definition the report uses, with its four key components. Winterberry Group recognizes “platforms that (1) are able to ingest and integrate customer data from multiple sources; (2) offer customer profile management; (3) support ‘real-time’ customer segmentation; and (4) make customer data accessible to other systems.” The alarm sounds when taking into account just the small percentage of platforms that really make the cut.
“I didn’t think at first that this would be a headline,” said Harrison. “At the beginning of the research in November, we started with over 100 vendors. When we get down to it, 20 or less are CDPs.”
“Also, I would have thought at the beginning that the CDP was replacing something in the stack,” he added. “In almost every instance, a CDP is additive to the [marketing tech] stack. I think that what we’re seeing is that the CDPs of the 20 that are really focused on mid-market are adding applications. The mid-market doesn’t have the staff to put it all together, and they would like a one-stop-shop…Out of the 20, at least 10 are doing that heavy [data] lifting and then doing the execution. The CDP is one component of their offering, but not the offering.”
It makes sense that vendors want to put together essentially “mini-martech stacks,” in Harrison’s phrasing, and package that as a CDP. But, he explained, “CDPs do measurement and attribution of campaigns…as opposed to execution. They are more natural to the data layer than the application. [layer]” Anything beyond the core data ingestion, management and accessibility shouldn’t be judged a CDP.
Another good question to ask about a CDP offering: how real-time is it? “It is not necessarily new and can be handled in alternative applications, but when speaking to marketers and platforms, the ability to segment and activate the data in real-time across the board was critical. Which makes sense when you think of the use cases of personalization,” Harrison said, elaborating that in such a case, for instance, when a customer comes to a site and the data requires actions, “layering in another application delays the response.”
That’s why the CDP has to be all-inclusive. But there is also a sense from the Winterberry Group report that the CDP may yet be a temporary solution, soon to be eaten up by other technology currently on the scene. One likely suspect from the report is the Master Data Management (MDM) solution.
“MDMs are critical in highly regulated industries such as financial services and health care,” Harrison said. “The MDM is required for the PII data, and the CDP leverages the data from the MDM and digital sources to create a ‘universal view of the customer.’ We have already seen the classic MDMs acquiring CDPs with the acquisition of Allsight by Informatica.”
Further consolidation across the CDP landscape is also expected by the team’s research.
“Winterberry Group believes that the CDPs will have to evolve and offer additional applications – i.e. machine learning, decisioning [and/or] orchestration, or personalization – or be absorbed by the larger martec stacks,” Harrison stated. “With over 100 CDPs of all flavors, we believe there will be consolidation of the CDPs over the next couple years.”