Customer Data Platforms have been widely adopted by retail marketers today as they attempt to get to grips with the ever-developing marketing landscape. However, as they are widely seen as an essential part of a marketer’s tech stack, many vendors have labelled their products as CDPs even though they are each providing similar but subtly different platforms.
We wanted to take a look at the CDP and explain how they have evolved to where they are today, and what problems they are intended to solve.
At the heart of a CDP is customer data. Before it was termed “data” it was more commonly called “information”, and stores of customer information predate the digital era. Think of things such as Rolodexes or address books, all of which are analog ancestors of the modern CDP.
But the real first steps on the road to CDP were when these Rolodexes were digitized in the late 1970s when people started putting the data into spreadsheets and writing macros to look up a customer’s address or phone number.
By the late 1980s and early 1990s this had evolved into basic customer relationship managers (CRMs), which still exist today. In those days they were intended to be the domain of customer service professionals, logging information about specific customers and prospects and what they knew about them. What products they had bought, any complaints or customer service issues that had arisen, etc.
By the turn of the millennium, these CRMs were more widespread, and the next change was to move into the cloud. Maintaining these CRMs on-premise was a headache for a lot of organizations as they would have to update the software regularly across an entire system. Salesforce was among the first to pioneer and promote the idea of cloud-based software-as-a-service and to turn the CRM into an online solution.
At the same time, the Internet age brought about online advertising and for marketers, this generated a whole bunch of anonymous data that they had to manage in order to create audience segments for advertising. As a result, Data Management Platforms (DMPs) were created. Whereas CRMs were about creating and logging one-to-one information about individual customers, DMPs were about managing a mass of data from 2nd and 3rd party software, such as these advertising networks.
Both of these systems evolved with the general technology landscape, but it was in 2013 that the CDP was first defined and developed. Roughly speaking, CDPs synthesize the personalized information from a CRM, and combine it with the masses of data generated by advertising and online activity, as well as all the data contained within marketing platforms such as email service providers. The idea was to create a central place to see all possible data about your customers.
As ecommerce has grown, and more marketing channels have been created, there is now an unmanageable quantity of data. Previously, each way that a customer might interact with a brand was tracked and that data was kept in the platform that controlled that channel. For instance, any interactions with Facebook advertising were most likely tracked in Facebook.
For marketers to create engaging, relevant and personalized experiences, it was necessary to understand what was going on in each of these channels. This was practically impossible because it involved a huge range of data across different platforms that had to be aggregated, with data connected based on the individual customer. Plus, as customers were shopping all the time, this data would constantly be changing.
What marketers were looking for was a single customer view, or a 360-degree view. This means that all the customer data is in one place, and marketers could see all the data associated with an individual customer.
This is what CDPs were created to solve. A CDP is, the CDP Institute describes it: “a persistent, unified customer database that is accessible to other systems.” Crucially though, it is a marketing platform, first and foremost.
The idea of CDPs caught on, and by some estimates there were over 100 vendors by 2019, just 6 years after the term was coined. As often happens when a term becomes fashionable, it can get applied to solutions that are not strictly CDPs as the CDP institute defines them, and are actually closer to CRMs that have some automated data ingestion attached to them.
Naturally as more channels and technologies have been developed, the data generated by these have been incorporated into the CDP, creating an even bigger pool of data which needs to be managed.
An unintended consequence of this has been that many marketers end up using more than one CDP to manage different use cases because one is not enough. The result? In the pursuit of a single source of truth, retailers have actually created added additional layers of complexity and sometimes obscurity.
While it was clear that customer data had proliferated across multiple channels and systems, and there was a need to consolidate this data, arguably it has not been worth the time or effort.
We surveyed 300 CRM Marketers last year, and 36% complained about not being able to connect data across different platforms, 28% said that data was inaccessible to them, and 20% said they had no single customer view. So clearly there are still some problems with data management.
Perhaps more importantly, is it really necessary to collect all this data? Given that certain data is more relevant and more important than others, should it all be collected and treated the same? For example, it is surely more important to have data about previous purchases available than it is to have which online advert they last saw.
Let’s say you have all the data that you could possibly get. That’s an awful lot of data to manage, much of which may be completely unnecessary. But how can you know which is important and which isn’t unless you sift through it? Even then, the fast-moving nature of marketing means that the conclusions you draw could be out of date soon. How many times have you seen Cost Per Acquisition jump up or down in one channel for example?
Beyond that, once you have the data in one place, are you able to properly use it? We’ve explored some of the concerns about a standalone CDP in this blog, but the bottom line is that unless you can use this data to create joined-up marketing experiences then there’s not a lot of point in having the solution.
The CDP is still a useful tool if it is used right, and if it has the functionality needed for your purposes. Beyond that, we are likely to see much more AI capabilities being applied to it to find interesting customer segments and potential opportunities to improve marketing performance.
David Raab, founder of the CDP Institute, wrote an article in 2019 called “Why your CDP needs a brain”. The bottom line was that to do smart automation, there needs to be real-time customer data, and also the ability to make the right decisions with that data. Arguably most CDPs do the first part well, but most don’t get close to the second. So we’d expect to see more intelligence layers being applied to customer data.
With changes to third-party cookies, it’s possible that the CDP will become even more important as a store of first-party data. This may be the best way to effectively market to customers as some advertising channels may wane.
We’re also seeing the rise of the CDMP – the Customer Data and Marketing Platform. This evolution addresses many of the challenges that standalone CDPs can create, including the common disconnect between the CDP and Marketing orchestration tools, by bringing the end to end process closer together, and reducing marketer reliance on IT teams, increasing marketer autonomy and speeding up execution.
CDMPs use the idea that marketers need customer data within their marketing platforms in order to create the personalized experiences that their customers want and expect. So by bringing the data capabilities of the CDP into a platform that can build cross-channel experiences, everything a marketer needs is in one platform.
This makes the process for the marketing user leaner and more effective, while not compromising on the data side. Users don’t need to switch between systems, which makes the process smoother, and can access all relevant data from the CDMP.
To find out more about CDMPs, check out our CDMP buyer’s guide here.