By Mike Becker and Emery Waddell

Brief Background

An old problem is being drawn to the forefront by explosive growth in the marketing technology landscape, as inaccurate and incomplete customer data flows through increasingly powerful tech stacks.  Data quality has hampered marketing performance for decades, but it didn’t feel quite as bad putting low grade fuel into lower performance marketing processes.  Marketing technology engineers have been hard at work building a new high-end vehicle for marketers to drive, and it’s no surprise that data quality, transparency, and manageability have emerged as gating factors on performance.  A number of new vendors are emerging to address this, most of which could be considered various flavors of customer data platforms (CDP), a nascent but quickly growing category.  We are big proponents of the opportunity in this space, and have an investment in one of the emerging leaders: Synthio.

What is it?

According to the newly founded Customer Data Platform Institute:

A Customer Data Platform (CDP) is a marketer-managed system that creates a persistent, unified customer database that is accessible to other systems.

Broadly speaking, CDPs automate the collection, synthesis, storage and routing of marketing data, allowing sales & marketing teams to focus more on selling their products and less on data management and data quality assurance.

What does it do?

Provides a single, accurate view of the prospect/customer:

CDPs ingest data from a broad array of marketing applications within the enterprise (CRMs, marketing automation platforms, website form fills, etc.) to create an integrated profile for each prospect/customer linked by a single identifier.  Currently, most CDPs focus on prospect/customer information collected from direct touchpoints (1st party data), but some of the more sophisticated, data-intensive platforms can ingest, synthesize, and do longitudinal analysis to validate external (3rd party) data as well, thereby creating a proprietary picture of each prospect/customer that is both up-to-date and complete.

Connects to prospect/customer interaction systems:

Having a platform that aggregates customer data, validates for accuracy, and forms unified individual profiles in a single location allows for a database that is organized and accessible. Through either batch or real-time APIs, customer interaction tools can plug into the CDP and, in one stop, gain access to the most complete and up-to-date data available. The CDP effectively serves as the intelligent backbone enabling marketing and sales teams to leverage the full power of their marketing stack and create personalized experiences with each and every customer touchpoint.

Enables customer analytics:

Having a single source of customer data empowers marketers to quickly perform analytics that drive actionable intelligence.  Some CDPs have a built-in analytics layer while others integrate seamlessly with analytics-focused applications.  Key insights like customer habits, sentiment identification, conversion probabilities and many others can be attained when customer profiles are unified and complete.

B2B Use Case:

It should be noted that many CDP providers are focused on B2C (consumer marketer) use cases.  The distinction between B2C and B2B is important because the specific workflows and technical requirements of each are meaningfully different. 

For a simplified B2B example of the power of CDPs, consider the following all-to-common data scenario facing sales & marketing teams:

CDP Use Case FINAL

How does it do it?

CDP Model

The figure above illustrates how the data might flow once an enterprise has integrated a CDP into its marketing tech stack. The CDP communicates directly with the relevant data sources, intelligently maps the incoming data to the proper fields within each prospect/customer profile, applies appropriate overwriting logic (some data will be written over, some will create new fields, etc.), normalizes data fields, validates email addresses, and performs any other necessary tasks to maximize the utility of the data for the marketer. The newly minted dataset is then pushed as appropriate to outbound marketing tools, which in turn feed their results and updates back into the CDP. This circular flow of data allows the CDP to serve as a dynamic central repository that is maintained automatically.  Most CDPs provide cloud deployments of their platform, making it even easier to manage and scale customer data, and to integrate third party data sources and tools.

Why is it important?

CDPs address a severe pain point that has long plagued marketers: without a dynamic central repository, they are forced to carry out data management tasks that are both tedious and outside their areas of expertise, and rely on their overloaded IT departments for basic queries or data flow changes. Dealing with scattered customer data sources forces marketers and sales teams to bounce back and forth between systems, filling in gaps and resolving discrepancies manually or with rudimentary tools such as excel pivot tables. This wastes valuable time, reduces campaign results, and leads to inaccuracies that are frustrating for both the marketer and the customer.

CDPs help resolve these issues and provide marketers with the most accurate, up-to-date, and useful information available – ultimately allowing them to drive more efficient and effective marketing.

Why is it emerging now?

Roughly speaking, CDPs first emerged as a distinct category in 2013 although the term did not come into significant use until 2015. It began appearing in reports of industry analyst firms in 2016 and now it appears to be well-established in the industry lexicon. Key drivers of this timing:

  • The pain is more acute now than ever: The challenge of coordinating across data silos is nothing new, but the rise of “best of breed” MarTech stacks has amplified the problem for marketers, creating a whole new mess of data for them to grapple with. Opportunistic technology providers have taken notice, leading to a new crop of potential solutions.
  • The biggest obstacles have been technical: up until recently, cluttered and cumbersome enterprise IT infrastructure has inhibited the full realization of the CDP vision, though many other approaches have been tried. Master data warehouses and data lakes failed to address the pain points, largely due to rigidity and heavy ongoing reliance on IT.  DMPs have limitations (see below) and so do integration hubs: pulling in data from various apps on the fly typically requires IT expertise, leads to missing information like profile history and lacks persistent storage.

The combination of scalable cloud hosting, more sophisticated database technologies and more flexible IT infrastructure has spawned a new generation of solutions that actually address this increasingly severe pain point in the industry.

Where did these vendors come from?

Many vendors started with slightly different original products / businesses before arriving at their current CDP product.  Broadly speaking, most CDP providers can trace their roots to one of the following categories of technology: *

  • Tag management systems that first consolidated identifiers from Web pages and then expanded to include other online and offline data sources. These companies are typically strong at linking certain data types (e.g., website journeys) to customer profiles but may struggle to absorb and synthesize external data sources if they don’t come with clean, mappable identifiers.
  • Marketing automation / campaign management systems that used customer data to intelligently engage via web pages, emails, call centers, and other channels. These vendors have expanded their data capture and consolidation capabilities and made their databases accessible to other systems. These companies have strengths and challenges similar to those that evolved from tag management systems, including potentially incomplete scope of addressable data.
  • Data assembly systems (API hubs, proprietary data services, etc.) that were designed to make customer / audience data sets available to other systems. Some of these vendors’ CDP products were developed to solve broader data quality issues that have long existed in market. These vendors have effectively been CDPs in some shape or form from the start and likely have a head start with regard to capabilities as a result. Producing one superior data set by synthesizing multiple data sets is in these companies’ DNA.

*These categories generally align with those of the CDP Institute’s Industry Profile, January 2016.

How is this different than a DMP?

In simple terms, Data Management Platforms (DMPs) suck up, sort and house data for digital advertisers in a way that allows them to manage cookie IDs and other non-personal attributes to generate audience segments. There are a few particularly meaningful distinctions between DMPs and CDPs:

  • CDPs provide a central location for all customer data; DMPs are focused on ad targeting data and don’t easily share that data with other MarTech applications
  • DMPs do not typically deal in personally identifiable information (PII), only anonymized audience segments
  • DMPs are oriented around executing and monitoring discrete campaigns rather than managing data on a persistent, ongoing basis
  • DMPs are limited in their ability to address underlying data quality issues

DMPs are inherently equalizers, not differentiators. CDPs have the potential to create an advantage for companies by synthesizing the uniqueness of their 1st party data with the completeness of 3rd party data to create proprietary and complete individual profiles in one place.

Summary

CDPs are an exciting leap forward in the long march toward true one-to-one marketing.  While CDPs are not a magical cure-all and even the most sophisticated of the technologies still have maturing yet to do, the potential impact of these platforms is substantial.  Marketers have long dreamed of a world in which all customer and prospect data is standard, accessible and routable at scale. CDPs are demonstrating the power to bring that idea closer to reality.

________________

Additional reading: