Customer Data Platforms have grown into a billion dollar industry ($1.3 billion in 2020, to be exact, according to the CDP Institute) because they solve a pressing marketer need: easy access to unified customer data.
The need is driven by the explosion of disconnected customer data as new martech systems capture data across new channels and by customers’ expectations for unified, personalized treatment despite this fragmentation.
When this explosion began, marketers originally looked to existing systems for solutions but found that each failed for different reasons: CRM systems captured only some customer information; marketing clouds were themselves comprised of disconnected components; data warehouses were inflexible; data lakes were designed for data scientists.
It took some time for industry analysts and enterprise software developers to recognize this gap in the market, which opened the way for small firms to establish the category. Some of these firms were headed by marketers who had looked for a solution for their own use and decided to build their own when they found none was available.
Others were software vendors whose products addressed related problems, such as tag managers that were already expert at gathering customer data or personalization vendors that already assembled unified customer profiles for their own use.
By the mid 2010’s, the core goal of a CDP was clear: provide a single, unified data source to power analysis and consistent treatments across all interaction channels. This implied a set of core requirements.
A CDP needed to gather data from all sources, retain the original level of detail, assemble the data into unified customer profiles, store the profiles internally for as long as the customer needed them, and make its contents profiles available to any application. The CDP Institute codifies these in its RealCDP certification program.
Unfortunately, the very power of the CDP concept has led to continued confusion. The CDP supports activities ranging from customer journey analysis to ad audience selection to guiding call center agents. This variety makes it hard for potential users to get a clear picture of CDP’s value.
At the same time, marketers have so many other unmet needs that many look for a CDP that doesn’t just create customer profiles, but also provides tools to use that data for predictive modeling, personalization, campaign execution, and sometimes even message delivery.
Meanwhile, the popularity of CDP has attracted more entrants to the industry, notably including many companies already providing software that would feed data to a CDP or use the CDP outputs. These firms include email, mobile app, and Web site systems; marketing clouds, analytical tools and personalization engines; marketing data providers; and even operational systems such as ecommerce and loyalty products.
Each new vendor makes a passionate case for why their particular scope of CDP is the best solution for buyers’ needs, and vendors of all scopes make similar-sounding promises about the problems their system will solve. With more than 120 products now listed in the CDP Institute’s vendor directory, many potential buyers don’t know where to start in finding the right CDP system for their own purposes.
But buying a CDP need not be an impossibly complicated mission. You can easily whittle down the universe to a reasonable set of options with just three questions.
What scope of functions do I need? Am I looking to replace my entire suite of marketing tools, or do I already have satisfactory systems for things like email campaigns, Web site management, predictive modeling, and analytics?
Once you’re answered that question, you can focus on vendors with the scope of features that matches your requirements. (As with all rules, this one has exceptions: some marketers with demanding needs end up buying one CDP that is extremely good at building customer profiles and another that is stronger at applications such as personalization or message delivery.
At the other extreme, you might find a CDP that meets all your needs plus some features you don’t require. Both approaches can yield good results, although they probably will introduce more cost and complexity than finding one system whose scope closely matches you needs.
Which vendors are suitable business partners?
There’s no point to considering vendors who don’t sell their product in your region or can’t fit your budget. You might also want to look at vendors of a certain size – either small enough that you will be an important client, or large enough that you are comfortable choosing them as a long-term business partner. Some buyers will also want a locally-headquartered firm or need to ensure that their data will reside in a particular geographic location.
Industry expertise can also be an important screening factor: quite a few CDPs specialize in particular industries such as travel, financial services, or telecommunications. Those firms will have a much easier time understanding your data and will often have built special features to support industry requirements.
Again, there might be good reasons to look at options that fail some of these conditions, especially if your business is not typical of others in your industry. But it usually makes the most sense to start by looking for a fit along these dimensions and then expand your search if none of the candidates you find seems suitable.
What are my specific needs?
This is by far the most challenging question because you might not know the answer in advance. The way to resolve the uncertainty is to define the projects (“use cases”) you want to support.
These are often marketing campaigns but might also be more general goals such as creating unified profiles for a wide range of purposes; conducting a particular analysis such as customer journey mapping or lifetime value estimates; creating predictive models to select offers or recommend products; selecting campaign audiences for email, display ads, retargeting, or other purposes; managing real-time interactions such as Web site personalization or call center support; or orchestrating treatments for each customer across channels.
If you have a very long wish list, you’ll need to narrow it to specific use cases that you expect the CDP to make possible and exclude use cases that would require additional changes such as new delivery systems or organizational realignment. Then you’ll work through the steps needed to execute each use case, identify the steps you can’t currently complete, and figure out what capabilities the CDP must provide to close those gaps.
Those capabilities become the requirements that you look for in your vendor evaluation. (Note: you will find many vendor evaluation resources and a library of 50+ predefined use cases at the CDP Institute Web site at www.cdpinstitute.org.)
If this still feels too broad to be useful, here’s one more shortcut. Remember that we said you should focus on use cases that are only possible if you add a CDP? Well, in nearly all cases, those turn out to be cases that require sharing data across systems.
This might be a lifetime value calculation that combines online and in-store purchases, an email campaign selection that uses Web site behaviors to create audience segments, or a call center integration that shows what a caller has just done on the Web site. If you’re looking for a short, concrete answer to the question of “What’s a CDP use case?” you can start with that.
About the author
David M. Raab is founder of the Customer Data Platform Institute, which educates marketers and technologists about customer data management. Mr. Raab named the Customer Data Platform category in 2013. Mr. Raab is author of hundreds of articles on marketing technology and of the Marketing Performance Measurement Toolkit. He regularly speaks and teaches at events around the world. He is a graduate of Columbia University and Harvard Business School.