There are now more bytes of data in the world than there are stars in the universe. If that fact alone doesn’t make it startlingly clear, then nothing else will: Data is now among the prime currency for businesses. Truly, the stars are the limit for what’s possible with data today, from using it to create fully personalized customer experiences to improving reporting for more intelligent and predictable growth and everything in between.
But that data on its own isn’t what’s so valuable. Quite the opposite, disparate data can end up creating more of a headache than anything else. Unlocking the full power of those mountains of data to achieve any number of goals requires proper data management to unify, sync and activate data. Only when the data is properly managed and accessible to your team -- not to mention fully actionable -- will it live up to its potential.
To deliver on this imperative, most organizations turn to a myriad of tools. These tools can serve a variety of different needs throughout the full data lifecycle. Two of the most critical, and most often debated about, are customer data platforms (CDPs) and data warehouses. Many organizations find themselves debating which of these is better to implement. The truth is, the answer to the CDP vs. data warehouse debate very much depends on your business needs. Let’s take a look.
Understanding the Roles of a CDP vs. Data Warehouse
With so many tools for managing data that serve go-to-market teams now available, a lot of them have overlapping capabilities. So unless you truly know what you need for your business, it can be very confusing to differentiate among each of the tools.
As a result, determining whether a CDP or a data warehouse is right for your business starts with a clear understanding of exactly what each of these solutions offers.
The Role of a Data Warehouse
What it is: A data warehouse is a repository of data that aggregates different sources. It typically offers a snapshot of data at a certain point in time.
How it works: Data gets fed into a data warehouse on a “bulk loading” schedule from ETL (extract, transform, load) systems. This bulk loading schedule means that when you query data in a data warehouse, it’s never live -- there is always at least a slight delay.
Once data gets fed into the data warehouse, it lives in “fact tables” that are usually numerical and can reference other tables. Using a data warehouse involves joining these tables together with queries. This is a technical process that requires the skills of a data engineer.
How you use it: Data warehouses were primarily built for reporting and analytics purposes. They typically house information like sales order data, product catalog data or even behavioral events from users. Under this purpose, you might use a data warehouse to answer questions like: How many orders do we get per month for each type of product family?
Beyond this original purpose, there is a newer trend of using data warehouses to accomplish “reverse ETL,” a process in which you take data out of a data warehouse and feed it into go-to-market tools. This new use of a data warehouse accomplishes two important objectives:
- It allows companies to leverage existing infrastructure and other software investments in news ways.
- It brings the data into systems of action to allow for smarter business processes and market outreach. In this case, it gives business users faster and easier access to the data by bringing it directly to the systems they use, like a CRM or marketing automation platform (vs. having to get time from data engineers to query data)
Who owns it: Data engineers typically own a data warehouse, as they possess the right technical skills required to do things like join tables and run queries. They will then send the output of that data to BI or analytics tools (like Looker and Tableau) that provide business or data analysts an interactive experience to explore the data.
The Role of a CDP
What it is: A CDP unifies disparate online and offline data sources to create a single customer view. It is accessible to other systems and is typically managed directly by the marketing team. Importantly, a CDP is a live system that is also a system of action, unlike a data warehouse, which is a system of record.
How it works: Similar to data warehouses, CDPs also bring in data from multiple sources. However, a key difference between these two solutions is that CDPs then feed this data into a known data model that centers around the idea of a unique identity. Quite simply, the point of focus for a CDP is individual people and/or companies (aka the “who”).
Bringing data into a CDP requires a crucial step called identity resolution, wherein data gets associated with an “identity.” This identity can either be a person (at Hull we call this a User) or a company (at Hull we call this an Account). The data warehouse requirement of joining together tables becomes irrelevant in a CDP because of this identity resolution step. Taking away this technical process enables go-to-market teams, like marketing, to manage the CDP directly.
How you use it: Because CDPs are systems of action that have the flexibility to ingest live, real-time data as well as data that’s not real-time, the use cases vary tremendously. They can span everything from unifying data sources to managing and cleansing data to activating data for use in personalized campaigns. Some of the top use cases include lead scoring and qualification, data enrichment and personalization.
Most organizations feed data from a variety of sources, including data warehouses, into their CDP. Overall, the flexible nature of CDPs makes it easy to mix and match data from different types of systems to all converge (and then become standardized) within a CDP.
Who owns it: Marketing operations teams generally own the CDP, as it does not require any specialized technical skills. Having the CDP sit within a go-to-market team positions organizations well to improve their understanding of customers and deliver more informed outreach.
Identifying How CDPs and Data Warehouses Can Work Together
Understanding the different nuances of a CDP vs. data warehouse is an important first step in determining the right solutions for your business. Equally as important is understanding how the two might work together.
Data warehouses are extremely flexible when it comes to data modeling, which means they can handle basically every type of data. As a result, while they serve a very different purpose than a CDP, they can still be useful for many organizations -- even in conjunction with a CDP.
For example, you can take the “interesting bits” of data from your data warehouse and map them to specific identities within your CDP. This helps paint a more complete picture of the users and accounts within your CDP to power even more informed decision-making. Two important things to understand with this type of use case is that (1) the data warehouse will still require a data engineer and (2) the data feeding into your CDP from your data warehouse won’t be real-time.
Settling the CDP vs. Data Warehouse Debate for Your Business
Based on all of that, what if the CDP vs. data warehouse debate isn’t actually a debate at all?
Many organizations have made significant investments into data warehouses over the past 20 years. Continuing these investments has come into question more recently, in large part due to the non-real time nature of data housed in data warehouses and the requirement of the specialized technical skills of a data engineer to access that data. In today’s fast-moving world, you need real-time data and your go-to-market teams need to access it in any moment of need.
This situation does seem to tip the scales in the favor of a CDP, but maybe a CDP is actually a different, better way to leverage existing investments in a data warehouse. Specifically, a CDP can help activate the data to provide a better return on the initial investment in a data warehouse.
In fact, there are several use cases where CDPs and data warehouses really shine when working together. Take marketing attribution: A CDP can unify all touchpoints, activity and behavior from a variety of SaaS tools, databases and even data warehouses to provide a truly three-dimensional view of how different users or accounts engage with your business (whether that’s marketing outreach, products or anything else) over time. You can then push this information back out to a data warehouse to link it with a tool like Looker or Tableau for more advanced analytics and reporting.
Where Will You Land in the CDP vs. Data Warehouse Debate?
The CDP vs. data warehouse debate is an important conversation to have as your business thinks about new and better ways to tap into a variety of data. What you’ll find along the way is that, depending on your current setup and goals, it may not be an either/or debate after all.
However you slice it, it’s important to take the time to understand what each of these solutions does and the unique role they can play in your organization. That understanding should give you a clear picture of what’s right for your business and how the two can potentially even work together to make sure you squeeze the most value possible out of all your data.
Angela brings over a decade of B2B technology marketing experience to her role as the Director of Marketing at Hull. Prior to Hull, she spent 5 years at utility data aggregator, Urjanet, where she held various roles in demand generation, marketing operations, and product marketing.