Download this article as a PDF.
The Customer Data Platform (CDP) industry has moved from muddled definitions to a more structured industry with clearly defined types that have been bucketed based on capability and use cases they support.
There has been a lot of activity in this industry with established vendors strengthening their capabilities and positioning in the market and an influx of new players with point solutions. While the build vs. buy debate continues, the decision is more lucid with many looking to buy ready-to-deploy CDPs.
CDPs Have Evolved into Categories Owing to the Breadth and Depth of Functionality
Before we jump into what we must look for in a CDP to make the buying decision, let’s define the types of CDPs that are available in the market.
While CDPs started out with unifying customer data across systems, structuring the data for downstream analysis, they have evolved to become broader, integrated systems. CDPs enable data ingestion, segmentation and analytics, audience activation, and personalized omnichannel orchestration.
Based on the capabilities and functions, CDP Institute groups CDPs into the following categories:
These systems gather customer data from source systems, link data to customer identities, assemble unified customer profiles, and store the results in a database available to external systems. This is the minimum set of functions required to qualify as a CDP under the CDP Institute’s definition.
In practice, these systems can also extract audience segments and send them to external systems. Systems in this category often employ specialized technologies for data management and access. Some began as tag management or web analytics systems and retain considerable legacy business in those areas.
These systems provide the features of a data CDP plus analytical applications. The applications always include customer segmentation and sometimes extend to machine learning, predictive modeling, revenue attribution, and journey mapping. These systems often automate the distribution of data to other systems.
These systems provide data assembly, analytics, and customer treatments. What distinguishes treatments from segmentation is that treatments can be different for different individuals within a segment.
Treatments may be personalized messages, outbound marketing campaigns, real-time interactions, or product or content recommendations. These systems often include features to orchestrate customer treatments across channels.
These systems provide data assembly, analytics, customer treatments, and message delivery. Delivery may be through email, website, mobile apps, CRM, advertising, or several of these. Products in this category often started as delivery systems and added CDP functions to support advanced analytics, personalization, or multi-channel campaigns.
A Full-stack CDP Supports Data, Decisioning, and Delivery
A full-stack CDP caters to all use cases in a way that is specific to the industry. A CDP is a means to an end – with the end being contextually relevant engagement. A full-stack CDP would enable:
- Streaming ingestion of demographic, transactional, behavioral, known and unknown customer data from online and offline systems.
- Identity resolution by creating a single, 360-degree view of the customer and a golden record by duplicating and enriching the data.
- Audience discovery and management with granular segmentation and advanced customer insights powered by micro-segments, segment analysis, churn, propensity and lifetime value analyses to drive next-best actions, and measure ROI with campaign and journey analytics.
- Real-time audience activation to drive hyper-personalized, journey-based marketing orchestration across online and offline channels and connect with customers in the moment.
Choosing the Right CDP
There is no one-size-fits-all solution to finding the right CDP. Choosing the most appropriate CDP is not about how feature-rich the CDP is but more about the need and the use cases you as a marketer are looking to address. Some of the key criteria to consider are:
Business Use Case Support
Clarity of what problem you are looking to address or what gap you’re looking to fill is crucial to choosing the right CDP. This translates to use cases. One of them could be to break data silos. You have customer data flowing in from various ingress systems, from various touchpoints. You have all this data, but it is not in a form that you could make use of to win customers over.
Another use case could be that you are not able to drive personalized engagement as you don’t possess deep insights on your customers. You don’t know what your customers’ affinities are, who is likely to churn, when or what offers would resonate with which customer. If you aren’t armed with these insights, your efforts to drive loyalty and improve basket size or visit frequency are unrealized.
Mapping and prioritizing use cases are critical to determine if you need a CDP and what kind of CDP you need. Most CDPs support data unification while many don’t enable analytics or activation. Hence, it is important to understand the capabilities provided by each of the CDPs that you are evaluating to determine if the use cases are supported.
There are many generic CDPs in the market that cater to both B2B and B2C businesses. The capabilities vary hugely. While B2C requires granular individual customer profile data, B2B is about lead management and account data.
Besides, the data structure and management needs of industry segments within B2C differ significantly too. The requirements of retail vs. banking vs. healthcare would be different. The AI layer that reads, understands, and analyzes this data needs to be trained on industry-specific data to be able to surface relevant decisioning intelligence.
The analytics models, domain measures, and metrics considered need to cater to grocery, fashion, or QSR industries specifically. Considering this, CDPs that are specialized to cater to industry segments deliver faster time to value.
Marketing Cloud vs. Best of Breed
CDP is not an independent layer but a part of a larger suite of products. CDPs act as superchargers to existing MarTech tools like personalization, marketing automation, and journey orchestration solutions.
The inflexible data management and profile unification features of marketing clouds were a major driver of marketer interest in CDPs from the beginning. These new modules aim to shift the integrated suite value proposition to a more open and flexible embrace of enterprise data, leveraging trusted relationships with CMOs and CIOs.
Solutions emphasize customer data management and connecting customer profile data to orchestration and execution tools within their products. Connections to technologies outside of the integrated suite for activation and execution vary greatly from vendor to vendor.
Data management has traditionally been the mainstay of the IT teams. But this is changing with end consumers of data increasingly showing interest in managing the data to leverage it for making informed decisions instantly.
Same is the case with marketing. As they own the budget for customer data systems, it only makes sense for them to own the end-to-end process of ingesting, managing, and activating this data for their specific use cases. The budget owner owns the ROI too, removing bottlenecks and making the entire customer engagement process seamless.
This brings us back to the usability of CDPs. Are they built for marketers?
Many CDPs were built clunky, making it impossible for marketers to use without depending on IT teams. As the need for marketers to be self-sufficient surfaced, CDPs are increasingly focusing on improving the ease-of-use so that marketers could use the various functionalities of CDP end-to-end.
This includes drag-and-drop data onboarding APIs, batch-loads and other methods, automated identity resolution, out-of-the-box analytics models eliminating the need for a Data Scientist to cull out segments or advanced analytics outputs, and out-of-the-box connectors to orchestration systems for seamless activation of audience for campaign purposes.
All of this is built with a high level of automation and intelligence making it easy for the marketers to use it. This, hence, is a key consideration in your purchase decision-making.
One of the core problems that marketers are looking to solve is unification of customer data that is sitting in various siloed systems. CDP is the go-to technology to address this issue by allowing direct data reading and automatic read extracts.
However, it is important to evaluate the connector ecosystem that the CDP has built to ensure you are able to drive seamless integration with systems of record that are already in place in your organization.
Automated data boarding, AI-powered data preparation, and schema-less data stores reduce the time and effort in CDP deployment.
Likewise, integration into egress systems such as personalization engine, marketing automation, or journey orchestration systems is important as well. These systems connect directly to the CDP or to the audience extracted from the CDP that is in a format that can be used for campaigns and other communications.
Hence, check for OOTB connectors to avoid custom integration. Know the data and the orchestration systems you need the CDP to connect so you are clear in your ask.
Start with Your Need Definition
The best place to start evaluating the need for CDP and what kind of CDP is by defining the detailed requirements. The next step is to chalk out the outcomes expected from the implementation of a CDP.
Download our CDP Checklist by Functionality & Use Case.
Use the above checklist for your convenience. Feel free to add to or modify it based on your specific requirements. You can then peg the CDPs that most closely fit your requirement.