Understanding the Customer Data Platform
Origins and the Different Business Models Defining the Category
2018 was a hot year for the Customer Data Platform (“CDP”) category. A number of marketing technology businesses have taken advantage of this new buzzword and either branded themselves as CDPs or associated themselves with “CDP-like” capabilities. With the number of CDPs hovering around 100 by some counts, the category has suffered from ill-defined specifics and boundaries around capabilities and outcomes.
The perspective shaping this article is informed by participating in dozens of evaluation processes across different business as well as consulting calls from industry analysts looking to better understand the CDP space. The key themes, patterns and questions that have emerged confirm that there’s no formal definition for what constitutes a CDP.
The CDP space is really an amalgam of different spaces, each with different origin stories and value propositions.
In order to help better understand the CDP landscape, it’s helpful to start with the key market trends underlying the industry and the resulting subcategories of CDP’s that have emerged as a result:
- DTC is King — Direct to Consumer brands are winning and are providing an elevated customer experience, raising the bar on what all brands need to deliver in order to remain competitive
- Scale and Flexibility Required — Organizations approaching an effective and modern cross-channel digital strategy have recognized the need for a data environment that is both channel-agnostic and able to handle significant data scale and complexity for diverse and unstructured data sets.
- Omnichannel → Customer Experience — Businesses can no longer strive to simply execute messaging in “every channel” as an end goal. Instead, successful marketers are considering their communication strategy with customers as a whole and consider the channels as a vehicle to deliver that communication effectively.
- The Walled Gardens are Winning — The dominance of Google, Amazon and Facebook’s identity graphs and the decline of the 3rd party cookie is driving a broad shift in targeting strategies. Advertisers are shifting their focus from buying cookie-based audiences on the open web to targeting individual customers.
- GDPR and Data Collection — Regulatory initiatives as well as consumer sentiment driving such regulations, have driven advertisers to focus more heavily on first-party data collection, and to be more strategic in its use.
In summary: (1) successful brands are developing a closer relationship with their customers, (2) that relationship is increasingly data driven, (3) the data requirements and sources to support these relationships are expanding in scope.
The CDP As A Solution to Evolving Needs
Some CDPs trace their origins to the dawn of the internet and digital advertising, while many are purpose built to address the emerging needs and capabilities addressed above. As a result, the CDP space is an amalgam of different technologies, with varied origin stories and value propositions colliding at the intersection of data and marketing.
As marketing technology systems have felt pressure to become increasingly data flexible (see: Salesforce acquisition of Datorama and Mulesoft, etc.), many non-CDP businesses have partially entered the CDP space or started marketing themselves as CDPs. Traditional marketing technology systems are fighting an uphill battle to build and / or acquire integrated CDP functionality for marketing.
These businesses can broadly be categorized into four buckets:
- Utility CDPs
- “Tag Manager” CDPs
- Marketing Clouds (not really a CDP)
- Marketing Orchestration CDPs
The Utility CDP
The Utility CDP was born out of the big data movement and coincides with the rise of “Data” as a function within enterprise organizational structures. Utility CDPs offer database storage solutions or sit on top of existing databases and power data utility workflows. These solutions may offer on-premise options and may not even have a user interface. Utility CDPs excel at customer record management, integration into legacy databases and many provide flexible “snap on” API connecters out into a host of systems as well. These systems are primarily built to de-duplicate customer records, create a marketing database or data store and handle the complexities of relational data required for marketing and analytics. Some of these solutions suffer limitations around use of real-time data for marketing use cases.
Utility CDPs are endeavoring to create more powerful marketing capabilities in an effort to move closer to revenue. As such, Utility CDPs are increasingly offering data science solutions and services and some are investing in marketing workflow capabilities. “Identity Resolution” is a newer term frequently repurposed for capabilities such as fuzzy matching and database cleanup as well as probabilistic cross-device targeting, which is an area in which some Utility CDPs excel.
- Companies with “old data” or those seeking to get value out of customer records in legacy databases
- Companies set on their marketing tech stack, looking for enhanced data capabilities
- IT teams or teams well resourced by IT
The Tag Manager
Tag managers were born during the rise of adtech, the desire to sync cookies and the host of associated web tags. Tag managers focused on the operational need to add and remove third-party pixels on a website as well as to sync event streams into marketing and advertising platforms. Over time, these companies worked to capture first-party authentication data and append it to CRM records, moving from the website into a more data-oriented architecture. These businesses are primarily focused on web and app data and anonymous customer records. Tag managers generally sell to technical and product teams who want to integrate data from one place to another quickly at a low cost. They excel at a narrow scope of functionality with massive scale.
Some tag managers offer or are beginning to offer segmentation and some basic marketing workflow. These businesses are building in both the data and marketing direction; they aim to increase their realm of available data beyond websites and apps, and also aim to capture value by offering a UI marketers can use to sync lists. Because these companies have massive install bases and are viewed as essential utilities for their product and technology teams, they stand a relatively strong chance of expanding their capabilities within existing clients. Tag Manager CDPs frequently exist in addition to the other types of CDPs discussed here.
- Broadly, everyone
- Businesses with lots of web / app traffic
- Technical teams / buyers
The Marketing Cloud
Marketing clouds were born in an era where sophistication (e.g. segmentation, testing, orchestration) lived in delivery systems (e.g. ESPs, DMPs, CRMs). As a result, SaaS juggernauts went on a buying spree of technologies they could build into all-in-one marketing clouds. Many businesses formerly branded as ESPs (especially those who lost market share as Salesforce, Oracle and the like won market share with their multi-purpose “clouds”) have branded themselves as “multi-channel messaging platforms” or even CDPs.
Separately, “new age marketing clouds” have emerged, touting their multi-channel capabilities and better user interfaces. Many of these businesses were born outside the email channel as app, push, or SMS targeting vendors and have since expanded their scope through bolt-on delivery integrations and white-labeling. These business are built to win in the marketing cloud space, basically for any business that doesn’t require a built in DMP (which is almost any non-media business)
The challenge with marketing clouds when it comes to capturing CDP market share is that their underlying data model struggles to handle the nuances and complexities of modern data environments (e.g. real-time, data science, relational data, scale, etc.). Further, they rely on APIs / SDKs, which puts the burden on IT teams to conform data to a specific schema and can’t scale well.
Marketing clouds are increasingly under pressure to deliver on the breadth of channels in which marketers want to engage their customers as well as the data demands of the modern enterprise. It’ll be interesting to see what happens to this space as businesses continue to focus on centralizing data capabilities and increasingly view message delivery (once the remit of marketing clouds) as commoditized (see: Twilio acquiring Sendgrid). Marketing clouds exist outside of the CDP realm and often exist alongside a Tag Manager CDP, a Utility CDP, or both.
- Enterprises / teams well supported by IT resources
- Businesses with advanced marketing orchestration requirements
- Businesses with existing data / CDP infrastructure supporting data requirements
The Marketing Orchestration CDP
Marketing Orchestration CDPs are born out of the inability of marketing clouds to deliver against marketers needs in the ever-changing data environment. In some cases, these businesses started as Utility CDPs and invested more heavily in marketing workflow or started as marketing clouds, but are architected on a different data infrastructure. They offer Utility CDP elements around consolidating data across businesses and generally focus on known customers. Their main limitations lie on both ends of the data to marketing spectrum: they don’t have all the features of a marketing cloud or a Utility CDP but offer 80% of middle of the spectrum.
Marketing Orchestration CDPs will usually integrate to a Tag Manager CDP for access to real-time information or anonymous customer data. These CDPs excel at marketing workflow and may even look mostly like using a fully-fledged marketing cloud. They may also offer predictive capabilities. Given their capabilities and offerings, Marketing Orchestration CDPs should be evaluated as a replacement to outdated CRM technology and as a bolt onto (or potentially even a replacement for) marketing cloud solutions.
- Marketers desiring data-rich environments for campaign orchestrations
- Teams looking to cut down on IT / engineering support burden
- Businesses with nuanced / complex data and use cases
The CDP vertical is fraught with buzzwords and feature creep — much like the broader marketing technology space where thousands of businesses claim to be a “true 1:1 omni-channel solution for marketers at scale.”
To better understand both where CDP companies fit within the space and which CDP is the best match for a given business — in addition to the high-level categorization above — teams should start with application, business value and use cases and work backward into technical requirements. This seems intuitive, but many teams are still trying to conform their CDP evaluation criteria into pre-existing requirements for CRMs, database applications and the like.
It will be interesting to see how the space evolves (and it certainly will expand) over the coming years.
Ultimately, there’s significant value to businesses in making investments relative to CDPs, but it’s important that the evaluation approach be outcomes driven and that teams are able to cut through the buzz.