Customer data platforms (CDPs) are still emerging as a solution class. The resulting fungibility from this status has led to extensive confusion in the marketplace, with some vendors claiming that the capabilities of a CDP are already fulfilled by other solutions. That isn’t true, but it hasn’t stopped the claims from being made. The emerging nature of customer data platforms also means they lack a market-standard definition, which allows some vendors to rebrand older solutions, such as tag management software, as a customer data platform.
The situation with CDPs is reminiscent of retail in the early 20th century, when the law of the land was “caveat emptor” or “buyer beware.” Companies who think they need a customer data platform – and many can, in fact, benefit from the solution – must be vigilant in evaluating vendor offerings. Many solutions branded as CDPs are limited in scope and functionality. This means that the vendor offering a customer data platform may not be selling brands the tool the organization thinks they’re getting.
In my experience, there are generally five key ways to tell if a solution is really a CDP. To be considered a customer data platform, a solution must:
- Accept data of any structure or cadence – Customer data platforms function as a single point of access and visibility for data silos throughout the organization. The high variance in structure and cadence of data throughout most companies – streaming unstructured data from social media, structured batch data from the CRM, etc. – means that any CDP worth the name needs to accept any structure or cadence of customer data. This is especially vital because Northeastern University recently found that 2.5 exabytes of data are produced every day. As volumes increase, CDPs need to have the ability to ingest broader varieties of information and still leverage it.
- Resolve customer identities across the anonymous-to-known lifecycle – Customer data platforms need to possess robust identity resolution capabilities. From a functional perspective, this means blending anonymous behavioral data with known information about customers into a coherent customer identity. With identity resolution capabilities, a CDP should be able to build and maintain a golden customer record from everything that is knowable about the average consumer.
- Manage customer data in real time – CDPs are designed to build and maintain the golden customer record. Any CDP worth the name must be able to perform the necessary tasks to accomplish that goal at the speed of the customer. While the tasks are often done at a real-time pace, the definition of “real time” is sometimes fungible based on business needs. Despite this, the fact remains that any CDP needs to be able to update the unified customer profile at a moment’s notice.
- Empower business users with access to unified customer data – The unified customer profile of a CDP is worthless if no one has access to that data. CDPs need to allow marketers and other business users to access customer records with minimal IT assistance; this is partly what makes CDPs so different from traditional data management technologies. Business users who can access customer data at their moment of need can do their jobs quicker, which leads to greater responsiveness to customer signals.
- Function on any deployment model and database technology – A customer data platform needs to be flexible in its deployment model. Some companies prefer on-premises deployments, others are fine with cloud-based software, and still others tend toward a hybrid deployment. The type of deployment shouldn’t matter. In terms of database technology, many brands have legacy databases in place that they don’t want to transition away from. A CDP shouldn’t force any organization to deploy a new database – the solution is designed to maximize investments in data technologies, not replace them.
Customer data platforms are a powerful solution class with a potent ability to ensure brands can deeply understand their customers. With a fuzzy definition in the marketplace, however, it’s crucial that companies know what to look for when evaluating solutions. If the solution being evaluated doesn’t accept all forms of data, handle real-time updating, make data accessible, have a flexible deployment model, or resolve identities across the anonymous-to-known customer lifecycle, then chances are the solution isn’t really a CDP.