Back to Blog

CDPs Are Not the Future

Tameem Iftikhar

Tameem Iftikhar

If you are a data-driven company, then you are probably familiar with CDPs. This is because in recent years, third-party customer data platforms (or "CDPs") have become a popular way to centralize and activate customer data. Alongside it, we have also seen the widespread adoption of data warehouses like Snowflake, BigQuery, or Redshift by tech-savvy companies. I'm here to talk about something a little controversial and make the claim that CDPs are not the future. For companies that have spent time & effort bringing all their data into a data warehouse, I propose a different approach, one that will streamline your ability to activate customer data & drive revenue faster—a First-Party Data Platform.

Your data warehouse should be your single source of truth and the place where you centralize, segment, and activate your customer data and a first-party data platform works directly off your data warehouse.

First-party data platform drives directly from the warehouse

So what is a CDP anyway?

Don’t get me wrong, customer data platforms provide tremendous value to organizations by centralizing customer data and activating it to the right business tools that teams use every day. This helps companies accelerate revenue and growth, and elevate the customer experience. Usually, CDPs do this by importing customer data into a single location, creating a single view of the customer, and activating the customer data to marketing, sales, and customer service tools. Just like any tool or platform, CDPs have their own limitations that we’ve come to realize over the years. And I want to run through some of these to back up the bold claims being made.

1. Time is Money 

In this new world where efficiency is everything, can you imagine 6+ months to set up and integrate a tool to activate customer data? That's the reality of using a Customer Data Platform. First, you will need to involve your engineering teams to integrate the SDKs/JS snippets and do new releases of your products. At mid sized organizations this process takes months. If you are a company that has a strong data engineering team that has already done the immense undertaking of designing a beautiful data warehouse with all your data—why would you now sync a subset of this data to a CDP?

Even apart from the data integration, you now also have to create and maintain your customer data model in the CDP platform. You will have to set up de-dupe logic, recreate your key customer attributes and create custom attributes you want to filter on (total spend, number of purchases, e.t.c). Most of the CDPs have their own nuances, so you will need to dedicate some resources from your team that will now be the "CDP admins" akin to something like a Salesforce admin. In a sense, CDPs force you to go all-in on their opinions and tools and dedicate your team to learning them.

Being obsessed with data-driven growth requires more flexibility and needs to allow for rapid experimentation—the ability to experiment with ML models, build analytical reports, or run advanced statistical evaluations on campaigns. That's why we see so many companies eventually bringing that integrated data from the CDP back into their own warehouse.

2. Two views of your customers

One of the main reasons for implementing a 3rd party CDP is to find a simplified way of creating a "single view” of your customers. Unfortunately, as you scale your company and therefore get into more complicated use cases, this can actually backfire and you will inevitably end up with not one, but two views of the customer- one in your data warehouse and one in the CDP.

Duplicate customer models are costly

Your analytics & data engineering teams are invested in ingesting all your data from disparate sources into the data warehouse and modeling it into a cohesive single view of the customer. This allows them to generate reliable analytics reports, build ML models, and do smarter evaluations on your marketing campaigns. These teams are also the most familiar with your key business logic and key metrics so there is no one who understands your data better. In the warehouse, your teams can leverage SQL and data modeling tools of their choice (Matillion, dbt, Informatica, etc.) to model data.

By bringing in the CDP you are maintaining a completely separate redundant data model with only a subset of the data that sits in your Data Warehouse. Over time, only the stronger data model will survive and almost always that ends up being the data warehouse. Analytics teams prefer to work with the tools they already love in the data warehouse, as opposed to the CDP, where they will be limited to using the proprietary query language or the user interface. Slowly but surely, all the work you have put into building and maintaining your CDP will cease to pay off.

3. CDPs have a "Limited View of the Customer"

It’s an unfortunate reality, but as the company scales, most of your customer data will never make it to the CDP. Even though CDPs can ingest copious amounts of data from various sources, it’s usually much easier to integrate simpler product metrics into CDPs such as — app events or website events. It’s a different story when you get to more complex first-party data.

Limited to easy sources of data

But the reality is never that simple, and you will have many more sources of data—survey data, email clicks, event attendance, sales activities, e.t.c. All of these will require extra integrations to sync to the CDP, and their customer models are not built to take this breadth of sources into account. Your data engineering & analytics teams, on the other hand, will be able to ingest and model this data in your data warehouse easily with tools they already enjoy using. So by depending on the data warehouse and activating it as the "1st party data platform" you will be able to activate on a much richer customer profile.

Sync all sources to data warehouse

4. Don’t sabotage your team’s ability to innovate

Analytics and data science teams work in an ecosystem of SQL/Python-based tools that are centered around data warehouse providers. By taking the data and putting it in a 3rd party CDP, you are handcuffing your analytics team and holding them back from innovating. Analytics & Data teams will come up with the best models for predicting churn, life-cycle value, and driving real customer value. However, CDPs make it harder for them to get the data they need to create these models and they will end up exporting the data back into the warehouse to be able to use the more powerful tools in their toolkit. A lot of CDPs don't necessarily make it easy to export this data back out of their system either because well, it's in their best interest.

Empower your data teams with tools they love

If you invest in your data warehouse instead, and build your core customer data model there, your analytics team already has all the data they need at their fingertips for analysis at all times. To put it simply, analytics teams prefer to use SQL & Python to 3rd party CDPs.

5. Data Warehouses offer more measurement potential 

With all the current changes across the industry in terms of attribution tracking for campaigns, it's more important than ever to have smarter audience/campaign evaluation mechanisms based on first-party data. Instead of trying to deal with attribution or looking at limited evaluation capabilities offered by CDPs, your data warehouse offers a much more powerful framework for reporting audience & campaign success metrics. Your data warehouse not only has a rich profile on your customers, but it also usually has all the target metrics or events you want to drive like total spend, purchase button clicked, and many others. So when it comes to analyzing audiences/segments your team creates for marketing and sales campaigns, it is a no-brainer to look at your data warehouse where you can tie those campaigns with all your events.

Empower your data teams with tools they love

By utilizing the data warehouse as a first-party data platform you are able to analyze audiences by looking at uplift in any metrics important to your business—retention, revenue, logins, pay now clicks, searches, quite literally anything available in your warehouse. You can track metrics like add-to-cart uplift, mobile login lifts, revenue lift by doing smart hold-out groups in the audiences generated in the warehouse. And you can run statistical significance on your results by using a Mann-Whitney U test. On top of this, your Business Intelligence team can visualize these results in your Data Visualization tool of choice: Tableau, PowerBI, Looker, etc. By keeping audience measurement within your data warehouse, teams gain flexibility and more insights, faster.

6. Don’t entrust key customer data to a third party 

Security is a big part of any infrastructure design or organization, and syncing all your data to a 3rd party CDP should definitely give you pause, and rightfully so. Some of the most glaring advantages of activating off your data warehouse are the security benefits that come along with this architecture design.

Security in your own hands
  • Your data never leaves - Your data stays within your warehouse
  • Access & Authorization - You control who gets access to which subset of data
  • Easy Compliance - You can control compliance directly in the warehouse
  • GDPR requests - Easily automate your compliance to answer GDPR requests to clear data

With privacy and security becoming an ever-growing concern on customer’s minds, you should never be risking their trust with a 3rd party.

A better way

Data teams are already heavily invested in building your data warehouse and data infrastructure. This means that there is a better, faster, and cheaper way to activate your first-party data directly from the warehouse. First-party data platforms enable your analytics team to leverage all the hard work they’ve already done and simply activate the customer data model you have in your data warehouse.

Customer segmentation platform on your warehouse

Full disclaimer, I believe in this new direction so much that we at GrowthLoop built an entire product focused on activating the data warehouse — a customer segmentation platform that runs directly on any data warehouse. We’ve helped clients like Indeed, Google, Red Sox, NASCAR, and others drive revenue using their warehouse. We are a part of this big transition towards a better approach to growth, experimentation, and marketing — unlocking huge potential for teams of all sizes and scales. And we are excited to continue discovering the potential of this new way of the future and pushing the boundaries.

Share on social media: 

More from the Blog

Press Releases
GrowthLoop Launches Computed Attributes generative AI feature powered by AWS to increase productivity and deliver hyper-personalized campaigns

GrowthLoop Launches Computed Attributes generative AI feature powered by AWS to increase productivity and deliver hyper-personalized campaigns

The solution allows marketers to execute highly-personalized, data-driven marketing campaigns in a fraction of the time with significantly less overhead. 

Cloud Data Warehouse
Guide: Using the GrowthLoop composable customer data platform on Amazon Redshift

Guide: Using the GrowthLoop composable customer data platform on Amazon Redshift

Discover how an AWS customer data platform (CDP) will help your marketing team create targeted campaigns faster, improving revenue and ROI.

Press Releases
GrowthLoop launches AI-powered audience generation as a Snowflake Native App

GrowthLoop launches AI-powered audience generation as a Snowflake Native App

Marketing teams can now quickly segment and activate their customer data directly in Snowflake, assisted by a generative AI engine that runs locally alongside their data.

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