Customer Data Platform: Definitions, tools and guide

Although the craze around “Customer Data Platforms” is probably a fashion statement, this dynamic in the customer relations market points to a need that has never been so important: the need for companies to rely on a 360° customer vision and to supply business tools (CRM, marketing automation, support, etc.) with reliable and unified data.

As we shall see, off-the-shelf CDPs are only a limited response to this need. Other approaches, such as customized CDPs or modern, “datawarehouse centric” CDPs, provide solutions that are better adapted to your business reality.

What is a Customer Data Platform? [Definition]

Customer Data Platform: Definition

The proliferation of acronyms in our business can sometimes lead to confusion, so let’s start by laying the foundations. The CDP Institute, a reference resource centre on the subject, defines the Customer Data Platform as “packaged software that creates a unified and continuous customer database that can be accessed by other systems and applications”.

Let’s decipher what’s behind these once again very jargonous notions: “packaged”, “unified and continuous customer database”, and “accessible by other systems and applications”:

A packaged software

CDPs are software, Saas, ready to use and to be integrated into your current data marketing system. The functional bricks are pre-built to facilitate rapid deployment and, above all, to allow business teams to get to grips with the system quickly.

This is a crucial dimension of the CDP value proposition: enabling business teams – marketing of course, but not only, sales and customer support teams – to integrate intelligence into their daily activities independently of IT.

https://images.viblo.asia/b061d5d5-fb12-4a50-8597-8a1f0c081e0a.jpg

Source: CDP Institute

A unified and up-to-date customer database

CDPs address the issue of customer data reconciliation. The race towards the ultimate customer relationship has led organizations to build more complex customer journeys, which rely on an increasing number of channels and tools.

Each of these tools generates data about each of your customers… to understand your customers’ behavior and track the many interactions with your brand, a “single repository” is essential. Simply put, it is a holistic view that links all the information about a given customer. This 360° view of your customers allows you to control the marketing pressure and of course to refine your messages a little more.

Your customer data is therefore centralized, standardised, and continuously enriched. By synchronizing the data mentioned in real-time (or almost), a CDP gives you access to an exhaustive, up-to-date view of your customers.

A platform that can be accessed by other systems and applications

A Customer Data Platform doesn’t just ingest data. It restructures it, adds calculated fields such as trends or scoring models, and then makes it usable by your other tools and systems through segments or aggregates.

https://www.customerlabs.com/wp-content/uploads/2019/12/guide-image-4.png

Source: CustomerLabs.

The diagram above illustrates the central place of a CDP within the data stack of organizations. It “unifies” data from different sources (on the left) and makes it usable by other applications, the destinations (on the right).

Finally, the great strength of CDPs, which distinguishes them from other data management tools such as CRMs, ERPs, or DMPs, lies in their ability to process all types of customer data: both “hot” and “cold” data, online and offline.

https://www.kameleoon.com/sites/default/files/inline-images/Thermometer-EN-min.png

Source: Kameleoon

Thus, a Customer Data Platform can ingest behavioral data, linked to browsing for example, and associate it with cold data: the information on your customers that you have in your CRM.

Customer Data Platforms: A fad that hides a real trend

The problem of reconciling customer data is not new. The most experienced among us remember that the promise of organizing and unifying customer data was also the promise of other families of tools, notably Marketing CRM.

/wp-content/uploads/2022/02/CDP-trend.png

The significant increase in search volume for the query “Customer Data Platform” since 2017.

Judging by the evolution of interest in the query “Customer Data Platform “ since 2014, this new family of tools is undeniably in vogue, and marketers are not mistaken! More and more software publishers are seizing on this vendor term for solutions that do not always correspond to the criteria previously stated.

However, behind this fashion effect lies a real underlying movement or at least a very clear acceleration of already existing needs:

  • The race for the “ultimate customer relationship”. Customer journeys are becoming increasingly complex, and all brands want to offer an omnichannel customer relationship (mobile application, automated chat, marketing automation, advertising retargeting, customer service, etc.).
  • The creation of data silos. Each of the channels mentioned generates ever-larger volumes of data, unfortunately, impossible to exploit when they are dispersed and “blocked” within these different applications.
  • The constant struggle between the creativity of marketers and the possibilities of CIOs / data teams. Marketing teams are yearning for more autonomy and are tired of being dependent on the data team to add intelligence to their campaigns.

The ability of the Customer Data Platform to provide a concrete response to all of these issues explains the growing popularity of these tools. As we mentioned, the interest in this new family of solutions has led many publishers to categorize themselves as such. As a result, tools of sometimes very different natures in terms of technologies are defined as CDPs, and we will decipher them in the following section.

Which approaches to building a CDP?

When it comes to building your Customer Data Platform, you have several choices:

  • Turn to an “off-the-shelf” or ready-to-use CDP
  • Build a custom CDP or customer database yourself
  • Opt for a hybrid approach with your cloud data warehouse as the operational foundation

Rest assured, we have put together this guide to provide you with some answers 🙂

#1 CDP on the shelf

This first approach brings together two categories of players: on the one hand, the new CDP offers developed by the CRM giants, such as Salesforce or Adobe, and on the other, the “pure players” in the market, such as Segment.

How could the giants of the Martech industry have missed such an opportunity? They couldn’t! At the end of 2018, Adobe, Salesforce, Oracle, and SAP all announced that they were developing a CDP offering.

Unification of customer data, 360° customer vision, tracking of hot data, and activation via your business tools, we find in the Adobe CDP offer all the essentials of the added value of Customer Data Platforms.

/wp-content/uploads/2022/02/Adobe-CDP-1.png

This is of course a very credible solution, but it has two structural limitations:

  • The astronomical cost of these solutions, which are inaccessible to SMEs / SMIs
  • These tools were not designed as open platforms but as closed software bricks integrated into the CRM ecosystem of each of these publishers, a major obstacle in terms of data control.

Mainly since 2015, many other players have also positioned themselves in this promising market. Often at the origin of Data Management Platforms or Tag Management tools, the number of these solutions has exploded in recent years. In its July 2021 report, the CDP Institute counted no less than 151.

It is difficult to provide a common definition for all of these tools, whose functional scope is largely influenced by their heritage: old Tag Management tools, web tracking, Analytics, or even DMP.

However, there is a common problem with all these tools and it is inherent to their value proposition: these ready-to-use solutions provide you with preconceived data and scoring models.

Unfortunately, if you want to move away from these models you will have to wait for your vendor to evolve its product. This is an important limitation in terms of flexibility, a question to ask yourself if your business model is not based on the traditional models.

#2 Build your CDP

Aware of the structural limits of the off-the-shelf CDPs we have just mentioned, many CIOs are choosing to build a custom CDP, adapted to their needs, which is a real industry trend.

This approach has the following advantages:

  • Off-the-shelf CDPs limit the number of use cases simply because there are as many as there are business realities: the possibilities are almost unlimited. This option allows you to focus only on the use cases that matter to you, and develop new ones as you grow.
  • Data control – off-the-shelf CDP vendors store your customer data on proprietary servers and don’t always make it easy for you to access it. Building your CDP means taking back control of your data by hosting it on your servers.

Of course, this is not a miracle solution. Building your own Customer Data Platform requires resources, skills, and time that few organizations have. It takes a minimum of 2 to 4 months to build an in-house CDP, and this is done by employing an army of data engineers.

Ultimately, building an in-house CDP is a solution only considered by large companies that are mature in terms of customer data exploitation.

#3 The hybrid approach: making your cloud data warehouse your operational base

Customer Data Platforms on the shelf very often claim to be able to become the single source of truth for your data system. In reality, for many companies, this single source of truth already exists, it is your cloud or modern data warehouse.

Until 2012, when Amazon launched Redshift, a data warehouse served as a base for analytics only and was only designed to support ad hoc queries, which occurred once a day at best.

Modern data warehouses (Snowflake, Google BigQuery, Firebolt) have broken through this glass ceiling by increasing computing power exponentially, while drastically reducing storage costs.

Our friends at Castor have written an excellent article on the subject that we invite you to read if you are interested in the subject.

This major technological breakthrough has led to the development of a whole ecosystem in which the datawarehouse is the pivotal point, and we are talking more and more about the modern data stack.

/wp-content/uploads/2022/02/stack-data-moderne.png

This is all very nice, but what about my CDP? You’ll understand 🙂

Making your datawarehouse the operational customer repository of your data device allows you to meet the traditional use cases of a Customer Data Platform:

  • A unified customer view that gives you access to all their interactions with your brand, in real time and across all channels.
  • Integration of your advanced segmentations and scoring models within your business tools.

Without connectors, how will my data flow from my data warehouse to my operational tools? Indeed, a datawarehouse alone is not enough, unless you have the manpower in terms of data engineers to build and maintain the necessary customised connectors.

The solution is a new category of tool adapted to this precise need, we speak of CDP 2.0 or modern CDP which runs on an independent base which can be a datawarehouse.

/wp-content/uploads/2022/02/evolution-customer-data-1.png

These solutions meet the criteria that allow them to be categorized as CDPs, eliminating the pains mentioned for the first two solutions:

  • A single repository for all your data, not just your customer data, and in many cases this already exists: this drastically reduces costs and deployment time.
  • Reduced data integration and processing efforts. Your data is imported into one place and transformed once and then used in the rest of your stack later on.
  • Data control: You are in control of the data stored in your data warehouse and can choose to use that data in the operational tools that make sense to you, without delay.

Much lighter in terms of deployment and investment, these CDPs give SMEs the weapons to fight against large corporations or start-ups that rely on big fundraising by allowing them to become truly data driven, to conclude with yet another jargonous Anglicism 🙂