How Publishers Can Use a CDP to Manage, Analyze & Act on Their Data

One way to approach the topic is to look at what different types of CDP implementations make possible. Broadly speaking, CDPs can be placed into three categories. Some CDPs only build the unified customer profiles, some build profiles and provide tools to analyze them; some provide profiles, analytical tools, and marketing programs such as email campaigns or personalized web site messages. So it makes sense to discuss which applications each type of system supports.

Available applications are also determined by the types of data fed into the CDP. Implementations often begin with a single feed, and then expand by adding other, similar feeds before expanding to include a range of disparate feeds. Every CDP can unify disparate feeds, or it wouldn’t be a CDP. But because many companies add new feeds incrementally, it’s again worth considering what’s possible in each situation.

Combining these two dimensions of three values each gives the sort of 3×3 matrix much beloved by consultants (as seen in the chart above). It would probably be possible to give each cell a cute little name, but we’ll ignore that opportunity. Instead, let’s dive directly into what’s possible in each situation. We’ll start with a single input source:

Applications for a CDP With a Single Data Source

1. Clean up your data.
   You might wonder what’s to unify if there’s just one data source, but plenty of systems contain duplicate records, not to mention data that’s outdated, incomplete, or just plain wrong. A CDP can clean up data from a single system to make it vastly more usable than it started out. It can also place that data in more convenient formats, compensate for changes in source systems, monitor trends, track changes using historical data which might be dropped from the source system, and allow easy access without querying the source system directly. The result is to provide greater value than you’d get if you just worked with the data in its original environment.

2. Map the customer journey on your site.
   Most CDPs can do basic profiling and segmentation. An increasing number include more advanced features such as predictive models based on machine learning. Some have specialized capabilities such as journey path analysis or content analytics. These can all be quite valuable even if they’re based on just one source of data, allowing publishers to identify new opportunities to monetize readers and deliver more relevant content.

3. Personalize marketing automation.
   Marketing programs might select messages based on specified events, behavior patterns, predefined sequences, model scores, content recommendations, or other factors. The messages can also be personalized with data the CDP has cleaned and reformatted. The messages might be delivered in email, on a web page, mobile apps, direct mail, display ads, call centers, or other channels. Note that even a system with a single data source can push messages to multiple channels. For example, it’s quite common to for web behaviors to trigger emails and ad retargeting. Some CDPs do the actual delivery themselves; others connect with external systems such as email engines, web content management, or ad servers. Being able to run marketing programs within the CDP saves marketers from integrating other marketing systems, simplifying deployment, and reducing costs. In fact, the value created by CDP-based marketing programs is often the main economic justification for buying the CDP.

Applications for a CDP With Similar Data Sources

CDPs might get data from similar systems if a company has multiple web sites on different platforms, multiple products on different fulfillment systems, or separate systems for separate geographic regions. CDPs in these situations support all the applications available when there’s a single source, plus new ones including:

1. Standardize data across different products.
   The CDP can standardize identifiers from different sources, such as names and addresses, making it easier to build unified profiles. The CDP can also standardize data such as product IDs and extract common features such as content themes or entities, allowing more accurate aggregation and analysis. Data from similar sources provides more inputs to a “golden record” with the best information on each customer.

2. Understand the customer journey across your properties.
   The CDP can now create a broader set of analyses, such as behavior across all web sites or all mobile apps. Attribution analysis becomes significantly more meaningful as it includes behaviors across multiple sources, even if these are all in the same channel. Ingesting similar sources also lets the CDP enhance its contents with second and third party data, such as behaviors on other web sites, allowing even richer analyses, segmentations, and insights.

3. Identify opportunities to market more products to your audience.
   Data from similar systems lets marketing programs identify opportunities such as cross-selling and react to behaviors such as registration for multiple properties. Richer data can drive programs based on lifetime value or on behavior changes that indicate higher churn risk or new interests.

Applications for a CDP with Disparate Data Sources

Combining data from disparate systems is the ultimate CDP application. Of course, there are gradations within this category, such as having different online sources but not offline sources. Still, it should be clear by now that a CDP can be useful even if it doesn’t include every company system. New applications made possible by disparate data sources include:

1. Create a 360-degree view of your audience.
   The CDP can now produce true cross-channel profiles, giving something much closer to the ideal complete customer view. It can now store data that has no home within company systems, such as purchased lists, cross-device identifiers, and feeds from external sources.

2. Understand the complete journey of the reader online.
   The Holy Grail of multi-channel attribution is now, at least theoretically, within reach. So are cross-channel journey mapping and deeper understanding of content consumption. With a CDP that can analyze multiple first- and third-party data sources, machine learning and predictive models can work with more complete data and produce cross-channel recommendations.

3. Unify your marketing strategy across different channels.
   The complete cross-channel view enables coordinated cross-channel marketing programs. Specific features include picking the right channel for each message based on customer preferences, capping communication frequency across channels, and responding in one channel to behaviors in a different channel.

These are just some examples of the possibilities. The vendor-agnostic, educational organization that I founded, The CDP Institute, provides a number of case studies that explore other applications. It’s also important to point out that media companies with adequate analytical and marketing program systems in place can get these same benefits from a CDP that only provides data unification. CDP vendors incorporate analytical and marketing program features to help companies that need such systems, not to force companies to replace them. Remember, allowing easy access by external products is a core part of the CDP definition.