Customer data platform (CDP)

GrowthLoop Icon
Researched by
GrowthLoop Editorial Team
verified by
David Joosten

Key Takeaways:

  • Customer data platforms (CDPs) include components that store and organize data, build audiences, and sync audiences across various channels.
  • CDPs can pull data from a wide variety of sources, including online and offline channels.
  • There are two primary types of CDPs: traditional (or packaged) CDPs and composable CDPs. A traditional or packaged CDP is a standalone platform that stores your customer data while composable CDPs sit on your existing data warehouse.
  • CDPs help facilitate faster marketing execution, fewer data bottlenecks, and a more unified customer experience across marketing channels.

Table of Contents

What is a customer data platform (CDP)? 

A customer data platform (CDP) stores and organizes customer data from multiple sources. It is a centralized location for customer data, which helps teams manage a unified customer profile (also called a 360-degree view of the customer). With a complete view of the customer base, marketing, sales, and customer service teams can better understand customer behavior and create more tailored and effective campaigns. 

CDPs include:

  • First-party data - Data collected directly from your customers from online or offline channels, such as through a website form or product purchase.
  • Second-party data - A referral source or a business partner’s first-party customer data that they supply to you. 
  • Third-party data - Customer data collected by a third-party source with no direct relationship with the customer. This can include data collected through website cookies.

A diagram illustrating how CDPs collect and store customer data.
A CDP is a centralized location for customer data, helping teams manage a unified customer profile.

Primary CDP use cases 

Here are a few common use cases for CDPs:

A checklist of the four primary use cases for CDPs.
Primary CDP use cases

  • Cross-sell and upsell existing customers - According to a 2021 report from the CDP Institute, 65% of CDP users surveyed said customer value is the primary goal for their CDP. With a CDP, you have a central location for customer data that includes behavior across marketing and sales channels as well as your website or product. This comprehensive data helps you target customers for additional purchases. You can also use data to build predictive models that find a set of customers with the potential for high lifetime value.
  • Outbound marketing campaigns (acquisition) - Although creating a unified customer view is often the primary use case for a CDP, marketers also leverage the CDP’s ability to help with outbound marketing campaigns. CDPs can assist with customer segmentation, generating audiences, and pushing those audience segments out to different destinations. In many cases, CDPs can also provide data on real-time interactions, which comes back into the CDP data storage and further informs your customer profile.
  • Cross-channel journey orchestration - CDPs give marketers the power to create customer and prospect journeys across various destinations. A journey can include email, social media, search ads, push notifications, and more.
  • Churn win-back programs - In the same way CDPs can help you create prospect segments for personalized marketing campaigns, they can also help you develop segments for churned customers. For example, you can use data to develop an audience of high-value customers (high spenders) who have recently churned. Then, create an offer that is targeted at these customers with the goal of winning them back with another purchase.

Customer data platform examples

Some of the top CDP examples include:

  • Salesforce Customer Data Platform
  • Segment
  • Tealium
  • GrowthLoop* (*composable CDP)
  • Zeotap
  • Bloomreach
  • Microsoft Dynamics 365
  • Hightouch
  • Oracle Unity
  • Treasure Data

Types of data CDPs collect

CDPs collect a variety of data points. Below are a few examples.

E-commerce:

  • Online orders and transactions
  • Returns
  • Abandoned carts
  • Products viewed

Digital marketing:

  • Website and mobile behavior
  • Pages visited per session
  • Average session duration

Product:

  • Profile data
  • Product usage
  • Billing information

Supply/inventory:

  • Online pricing
  • Inventory levels

CDP vs. CRM

Like CDPs, customer relationship management (CRM) platforms store customer data in a central location, giving a full view of the customer journey from sales through purchase and advocacy. CRMs are most often used by sales and customer success teams for 1:1 prospect and customer communication. 

For marketers, some CRMs offer audience building and customer segmentation based on the data in the CRM. The data can be deployed with email automation tools in the CRM. 

But CRMs typically have a customer view that’s limited to sales and marketing channels built within the CRM (email opens, sales calls, etc.). CRMs focus only on identifiable customer data like email addresses, phone numbers, and locations.

On the other hand, CDPs can include data from a broader range of sources, including anonymous data from mobile devices or behavioral data from your product. It collects information beyond the customer’s sales, marketing, and customer service journey.

CDP vs. DMP 

While CDPs can create a more specific and targeted customer profile, data management platforms (DMPs) focus on anonymous third-party data to help build aggregate audience models. DMPs also store the data for a short time.  

DMPs are often used in advertising campaigns that target broad or lookalike audiences. 

It’s important to note that both Google and Apple have announced that they’re phasing out third-party cookie tracking, which will affect use of DMPs.

Types of CDPs

While traditional CDP software has been a part of the martech landscape for quite a few years, the composable CDP is a newer concept in the industry. The composable CDP is built on the data cloud, which provides analytics for audience segmentation and machine learning for customer modeling (propensity modeling). 

Traditional CDP vs. composable CDP 

The primary difference between a CDP and a composable CDP is how the platform components are packaged.

A traditional or packaged CDP is a standalone platform that stores your customer data and helps you activate that data across various channels and destinations — email, push notifications, search ads, etc. Teams must use the data models and data sources that the packaged CDP supports. 

While a composable CDP includes the same components as a CDP, it connects to your existing data warehouse and marketing channel tools. Because it’s built on existing resources, a composable CDP generates faster time-to-value and takes advantage of other cloud capabilities like machine learning. It sits on top of your data warehouse where your customer information lives, such as Snowflake or BigQuery. Then, it helps you create audiences with that data and use them on your existing marketing tools — such as a Hubspot CRM, Pardot email platform, or Google Ads.

With a composable CDP, you still control and store all your data — you’re not storing it in another third-party tool. Controlling your data is especially important for businesses with complex data and those in highly regulated industries such as financial services, healthcare, and education. You can also use the audience activation channels that you’re already employing in your strategy. This composable martech option allows you to scale and grow with the channels that meet your customer needs.

A comparison chart showing the difference between how packaged CDPs work vs. composable CDPs.
How packaged CDPs work vs. composable CDPs

How a customer data platform works

Customer data platforms include components that store and organize data, build audiences, and sync audiences across various channels. Here’s what the customer data platform architecture looks like:

Data storage and organization

A CDP starts with customer data. Traditional packaged CDPs store that data in the CDP itself. With a composable CDP, the data lives in your data warehouse — such as Snowflake or BigQuery. 

The CDP also helps organize customer data across different sources to create complete and accurate customer profiles (also called identity resolution). 

Audience and user journeys

Once the data is organized, CDPs make the data available for marketers to create audiences and map user journeys across a wide range of channels and destinations. 

  • An audience is a specific group or cohort of customers that you are targeting with messaging and offers. 
  • A user journey includes the different customer touchpoints you build for an audience. For example, a journey may start with an email, then a social media ad, then a push notification. 

Typically, a traditional CDP uses its own pre-built data points to organize customer information, and marketing teams must follow these parameters. For B2B marketers, this can pose issues with having a full view of account and contact data. And for B2C marketers with multiple products or product lines, a traditional CDP can make it challenging to store customers’ product-specific data.

Since a composable CDP sits on your data warehouse, there’s more flexibility with the customer data points you can use and track.

Audience destinations

Once you have audiences built and journeys mapped, a CDP allows you to activate those audiences to different destinations — email platforms, CRMs, social media ads, search ads, push notifications, etc. Packaged CDPs and marketing clouds typically prescribe specific audience destinations, while composable CDPs can use the platforms your team already has in place.

As the campaigns run, CDPs can gather data from the audience destinations and push that customer or prospect information back into the data cloud. This process allows your marketing team to iterate on and improve future campaigns. 

Why should I care about CDPs?

In today’s marketing landscape, it’s more challenging than ever to reach the right customers effectively. Why?

  • Most customers today expect personalized experiences and authentic interactions with brands.
  • Regulators have increased their focus on privacy, developing new laws, including the General Data Protection Regulation (GDPR) and the California Consumer Privacy Act (CCPA). 
  • Google and Apple have agreed to phase out all third-party data tracking (third-party cookies).
  • Marketers must keep up with the constant influx of new media channels, including TikTok, BeReal, Snapchat, and more. 

With these hurdles, you need to shift your focus to first-party data — data provided by the customer through an interaction with your brand. Then, you need to digest and organize this data into audiences and journeys that allow you to connect with customers meaningfully. CDPs can help with this. 

The benefits of a CDP

Whether you have a packaged CDP or a composable CDP, the benefits include:

  • Faster marketing execution speed - CDPs are built for marketers and are structured in a way that makes it easy for you to analyze customer data and build effective campaigns.
  • Ability to create a unified customer experience across channels (omnichannel experience) - Most CDPs have built-in connections with various audience destinations, such as Facebook ads, Google ads, CRMs, email platforms, and more.
  • Fewer data bottlenecks - A CDP will store all your customer and prospect data in one location. This approach avoids bottlenecks when trying to pull data from different sources. 
  • Data security and privacy compliance - Once your data is stored in the packaged CDP (or, in the case of a composable CDP, your data cloud), it tends to be secure. Managing a single customer profile (or unified profile) is also a more effective way of meeting privacy regulations.
  • Syncing data to a variety of channels - CDPs collect customer data (properties and events) and send that to multiple marketing and sales destinations, such as Google Ads, Facebook, email, etc.

The cons of a CDP

Although CDPs can unlock data for marketing teams, traditional or packaged CDPs have some drawbacks:

  • Time-consuming set up and integration - Setting up a traditional CDP requires engineering time and resources. Even for medium-sized organizations, this process can take six to 12 months. Composable CDP setup, on the other hand, can take just a few minutes.
  • Frequent code and integration updates - With traditional CDPs, the organization must constantly monitor and update the code to make sure the data is accurate. Without monitoring, companies may end up with hundreds of events and properties and inaccurate data.
  • Not all destinations can read the standard data library - While composable CDPs offer more flexibility for marketing destinations, traditional CDPs often lock you in to a limited set of channels. This means developers often need to go back and adjust code when a team wants to add new destinations.
  • Limited customer view - Packaged CDPs can consolidate your customer data from multiple channels, but it’s not able to collect all of the data about your customers. It can integrate simple data into CDPs, not more complex data. Each data source also requires a separate integration to sync. Composable CDPs, however, sit on top of the data cloud and can pull all data from that single source.
  • Copied data - Traditional CDPs require a data copy or transfer, which means its stored in a third party source and requires updates. With a composable CDP, your data never leaves the data warehouse.
  • Lack of AI tools - Because traditional CDPs copy and store data, they can't leverage any AI technology happening in the cloud data warehouse or data destinations. Instead, teams must rely on any AI innovations the CDP company develops.

Today’s CDP landscape

Customer data platforms have been around since the early-to-mid-2010s. These packaged, traditional CDPs evolved from marketing data cloud technology and customer relationship management (CRM) platforms that couldn’t keep up with the growing number of destinations marketing teams needed to reach. 

However, the last few years have ushered in an era with composable CDPs built on top of your data warehouse — a central location to store all customer data. A composable CDP is a more flexible platform that activates data from the cloud to your preferred audience channels. 

How many marketers use CDPs today?

A survey conducted by the CDP Institute showed that the number of marketers using CDPs has dropped slightly, but overall, more organizations have a CDP in place.

  • 20% of organizations surveyed in 2023 said they have a CDP in place. This number is up from 19% in 2022 and 14% in 2021. 
  • 25% of marketers surveyed in 2023 said they use a CDP, down from 31% in 2022. 
  • 33% of marketers surveyed in 2023 said they had a unified customer database, down from 38% the year before. 
  • 30% of organizations surveyed in 2023 said they plan to start implementing a CDP in the next year.

Why aren’t more marketers using CDPs?

Although CDPs provide many benefits from a customer journey standpoint, getting data into the CDPs can be a hurdle. With traditional or packaged CDPs, exporting data sources and importing them into the new CDP can be a long and complex process that typically requires engineering work. 

As data sources and volume grow, much of it goes into the data warehouse that data and engineering teams manage. This means most relevant data must be copied or transferred from the data warehouse to the CDP. Today, many marketing teams who need dedicated engineering resources struggle to maintain their CDPs.

Even composable CDPs require data in a data cloud or warehouse, and not all marketers have the resources to move data into these systems. Fortunately, there are solutions available that can help speed up getting customer and prospect data into the data cloud.

How many CDPs are available today?

The CDP industry has been expanding rapidly over the last decade. The International Data Corporation (IDC) predicts that the CDP market will grow to more than $5.7 billion by 2026. 

The list of CDPs on the market is extensive — the CDP Institute reported in July 2022 that there were over 160 defined vendors in the space. Some of the more well-known CDPs are Segment, Tealium, and mParticle. According to a search on the G2 software marketplace, there are 200+ listings for customer data platforms. 

How to build a customer data platform

Some teams may choose to build an in-house CDP rather than buying a packaged or composable platform. With more organizations adopting data clouds, marketers can work with analysts and engineers to activate data from the warehouse and launch campaigns.

Once you’re ready to activate data from your data warehouse, here are some critical steps to build a CDP in-house.

Hire a team of analysts

Marketers define their campaign audience to the analysts, who then write SQL queries to pull that audience out of the data warehouse. The analysts share comma-separated value (CSV) files across the team to review, and eventually, marketers launch the campaign with this data.

Build a marketing-friendly interface

To assist with data access and reduce the time to launch a campaign, some organizations build an interface that lets marketers pull data directly. This approach attempts to remove the need for a team of analysts by having the engineering team build a tool that allows the marketer to run their own SQL queries.

Assess how you will provide ongoing support

It’s also important to consider the resources your organization will need to support the CDP. 

Your marketing team will likely need access to new data tables in the data warehouse. Or, you will need to add a new end destination for your campaigns. 

When thinking about build versus buy options, consider how the marketing team’s internal product will receive ongoing support if you decide to build the CDP in-house.

Choosing the right CDP for your team

If your team is considering a CDP, it’s critical to consider your internal engineering resources and the current and future state of your customer data. Ultimately, you will want to map your business needs and use cases back to the CDP you choose. What requirements do you have for channels? For your data? Are there other capabilities you’re considering?

Key questions to ask

Be sure to discuss these responses and your growing needs with any CDP vendor:

  • Time to value: How long will it take our team to begin seeing value? How much engineering and data team support does the solution need?
  • Marketing execution speed: How self-serve is the new tool for marketers? 
  • Cross-channel campaigns: How well does the solution address the growing list of marketing channels? How easily can you launch cross-channel campaigns?
  • Data trust and reliability: How well does the solution tap into your company’s existing single source of truth for customer and prospect data?
  • Security and compliance: How important is information security to your organization? Will your CIO/CTO allow you to copy sensitive data into a third-party marketing system?
  • Engineering resources: Do we have the engineering support to implement this tool? 
  • Future-proofing: What is our maintenance plan if the tool requires our team to build a new customer profile within it? 
  • Standardized measurement: How well does the solution help the team standardize marketing performance measurement? Is it easily integrated into your organization’s Marketing Analytics?
  • Cost: How expensive is the solution? Does it fit in my organization’s budget? What can I foresee about the scalability and long-term ROI of this solution? 
  • Real-time customer data: How well does this approach serve real-time use cases like web personalization?
  • Data ownership: What if we want to change vendors one day? Should we risk building all our business definitions into a specific tool?
  • Artificial intelligence: How well does the solution future-proof your marketing to incorporate AI models easily?

How to read CDP reviews

Many marketers start with online reviews when browsing for a new platform. When looking through CDP reviews, here are a few critical areas to consider:

  • Implementation time and complexity - What do the reviewers say about how long it took to implement the CDP? Does this fit within your ideal timeframe? Did you need experts to help set it up? 
  • Engineering/tech support - What kind of support does the CDP provide throughout implementation and use of the platform? Does this align with your internal resources? 
  • User interface - How easy or difficult is it to use the platform? Do you need a technical background to get around the interface? 
  • Audience channels - Which audience channels does the CDP support? And how reliable is the platform in pushing audience data to those channels?
  • Industries - Does the CDP have experience supporting your industry? If not, how can they adapt to your specific business needs?
  • Data cloud compatibility - If you’re working with a composable CDP, which data clouds do they work with? Does your organization already use one of their compatible platforms?

Customer stories and industry validation

Not all CDPs may have public reviews available. When conducting research on CDP platforms, it’s also important to review customer case studies, references, and industry partners. Consider these questions:

  • Does the CDP work with other reputable industry leaders? 
  • Do they have case studies that match my industry or target customer? 
  • Are they able to provide references from current customers?

Phases of CDP adoption

If your team wants to implement a CDP, there are several steps you’ll need to take before you can launch your first campaign. Depending on the type of CDP you use, this can take from zero to three months (composable CDPs) to 12-18 months (packaged CDPs). 

For many teams, it helps to first define your use case for the CDP and work out implementation steps from there. For example, a B2B company may want to expand the number of engaged contacts per account by targeting active opportunities. B2C companies may want to reach target prospects on different marketing channels to boost engagement. The implementation steps would be different for each of these use cases, because there are different data sources and audience channels you’re targeting.

Once you decide your use case, there are five high-level steps of CDP adoption:

  1. Data strategy - Determine your current data sources and where your data will live with your CDP solution. Consider your data quality and complexity and how you’d like to use it to achieve your business goals. 
  2. Data ingestion - At this point, the CDP will start ingesting all your raw customer data. In the case of a composable CDP, this data goes into your data warehouse. Packaged CDPs will store the data in their platform.
  3. Data modeling - Now that your raw customer data is in the CDP or data warehouse, it’s time to organize it to make it useful for campaigns. This process includes identity resolution, or identifying the same users in different data sources.
  4. Data visualization and intelligence - With clean, organized data, you and your team can now dive into the analytics behind your customer data. Using analytics tools like Looker Studio or DataStudio, you can visualize important trends in customer data that help inform audience journeys and campaigns.
  5. Activation - With customer and prospect insights, you can create audiences and user journeys across different destinations. Then, export audiences to email platforms, CRMs, social media campaigns, search ads, etc.
Published On:
October 20, 2023
Updated On:
May 17, 2024
Read Time:
5 min
Want to learn more?
Contact Sales
NExt Article
Composable CDP
Previous Article
Generative marketing

Looking for guidance on your Data Warehouse?

Supercharge your favorite marketing and sales tools with intelligent customer audiences built in BigQuery, Snowflake, or Redshift.

Get Demo