For as long as there has been customer data, there’s been a need for data quality. Yet even though revenue often depends on having deep insight about a customer, a household, a product, a business, or another entity, data quality challenges persist across industries. Pain points like low customer satisfaction, churn, overmarketing, and taking on unnecessary risk all stem from poor data quality.
With so many investments in customer data technology intended to solve for the problem of low-quality data, it’s hard to fathom why the problem is still so pervasive. It’s certainly not for a lack of trying. Rather, many organizations simply aim at the wrong target. They try to solve the data problem in the IT department, such as with a master data management (MDM) solution. One pitfall with this approach, however, is that IT is generally not incented to tie the master data to the customer signal infrastructure. That is, IT will be a data steward, but is not concerned with the reasons why a master record is important for a marketer or a business user. The actions a customer takes (transactions, behaviors) and the actions a customer may take (intent, predicted behaviors) are outside the scope of IT. Clean, accurate data is not recognized as an enterprise principle – that there is an actual business or CX purpose behind the need for master data management.
Data Quality and a CDP
Bridging the gap between mastering data and attaching behavioral, transactional and predictions was the original intent behind a customer data platform (CDP). By overlapping with an MDM and layering in customer signals to a master record, a CDP would ideally present users with data ready for their CX use cases.
Instead, with a few notable exceptions, what transpired was that CDPs took upstream data, unified it somehow, and presented it to marketers without fixing the data quality issue. In most organizations, there is still a wide gap between master data and data that is ready for business use, with the output of an MDM and a CDP siloed to one another.
Efforts to bridge the gap, to overlay the understanding of what a customer is doing with a master data record, have fallen short.
A Stopgap Approach to Data Quality
One approach is to use AI to improve the output of identity resolution. The issue, though, is that while this and similar efforts (e.g., basic deterministic matching) may fulfill the promise of better matching, they still leave raw data in the system without any means to refine that raw data into meaningful, accurate, up-to-date and detailed customer experience data.
Another attempt at making data ready for business use has been to deploy a composable CDP on top of an existing cloud database, bringing the data closer to various endpoints. But this too has a few shortcomings. First, this method still runs on the assumption that data quality is being taken care of somewhere upstream. This has been the promise of the “marketing clouds,” which for the most part do not even claim to be responsible for data quality.
Businesses need to recognize (data readiness) as a foundational enterprise principle. A CDP alone cannot fix poor data quality, and traditional stopgap measures do not ensure the accuracy, completeness, or relevance that is needed to support simple to complex business or CX use cases.
Yet another attempt is that, with a composable CDP in place, there is a default reliance that there will be a bridging strategy between the customer data in the cloud and the operational data needed for the CDP, e.g., segment definitions, campaigns, output, etc., which might even be operating in a different cloud. By copying a subset of data and sharing it, this approach attempts to federate a system that is not naturally federated. What inevitably suffers is the accuracy and completeness of the view of the customer or asset that you’re trying to understand.
A New Approach to Data Quality
What these stopgap approaches all share in common is that they all lose sight of data quality as being mission critical for the enterprise. The goal becomes more about moving data around so that a CDP can do some analytics and segmentation before activating the data, but the importance of upfront data quality is lost in the shuffle.
This becomes evident when looking at how marketers typically interact with a CDP, such as using GenAI capabilities where questions or tasks are skewed to make things easier for marketers. Questions are usually in the vein of asking the CDP to define natural segments, or to provide answers on how to market to a certain segment, or to build a segment so that the marketer does not have to use SQL. These are all enormously beneficial, but rarely do marketers ask about, say, the validity of a certain data source, or what data is providing the latest understanding of a household’s dynamics. These and other questions that get to the essence of data quality are typically not being asked – or answered – in most CDPs, clear evidence that data quality is, if not an afterthought, generally thought of by marketers as being someone else’s responsibility.
Data Quality, an Enterprise Initiative
If organizations want a complete and reliable understanding of their customers or other enterprise assets, they must shift their approach to customer data technology. Rather than treating data readiness as an afterthought or placing the responsibility solely on IT, businesses need to recognize it as a foundational enterprise principle. A CDP alone cannot fix poor data quality, and traditional stopgap measures do not ensure the accuracy, completeness, or relevance that is needed to support simple to complex business or CX use cases.
The key to solving this persistent challenge is to prioritize data readiness – where high-quality, well-structured data serves as the foundation for analytics, segmentation, and activation. By doing so, companies can ensure that both IT and marketing work from a shared, trustworthy data source that drives better customer experiences and business outcomes.
As this series continues, we’ll explore how data readiness as an enterprise principle bridges the gap between raw data and actionable insights, and why it should be the foundation of a modern customer data strategy for the enterprise.
