The CDP graveyard: MarTech’s once promising solution lost touch of its purpose. Here’s how it’ll regain it.
The problems CDPs were always intended to solve
About six months ago, I wrote about how the CDP space would need to evolve in light of MarTech consolidation and headwinds in enterprise software. The conclusion was:
“To be relevant in the current MarTech gauntlet, a CDP must not only help you move data around, but it must also help you recognize your customer, help you know things about your customer that you don’t know today, and most importantly, it must do that in the service of measurable marketing ROI. “
My prediction then was that infrastructure CDPs — like Segment and Hightouch — would push to focus their product marketing on marketing ROI. I mentioned examples of previous times CDPs have made product and packaging decisions to get closer to the marketing persona, often with real challenges (e.g., Twilio Engage).
The “composable” trend has significantly changed the CDP category over the past several years. It has created much more cost- and resource-efficient marketing data operations. It has also shifted ownership of the CDP toward the IT or data persona and pulled the CDP further from the stakeholders and business outcomes it was intended to impact.
So, as the road ahead mandates clear incremental business value from a CDP investment, the category has, paradoxically, prioritized the “ways of doing the job” instead of the “jobs to be done.”
One big caveat is that with improvements in neural networks, data structure will become crucial to training and, therefore, AI-driven marketing campaign ideation, optimization, etc. This makes the end value a marketing team can create fundamentally a data problem.
This is the problem CDPs were always intended to solve.
The disconnect is that CDPs must bring these solutions to market with crucial context on the problems that marketing leaders and operators are trying to solve. Undertaking an initiative focused on the “ways of doing the job” (AI-driven marketing optimization, in this case) is a recipe for team misalignment, scope creep, and a result void of measurable outcomes.
Raise your hand if you’ve undertaken a technology investment in the past several years that failed to realize the intended business outcome. Was there strong alignment from the beginning on the outcome, and was it shared across teams?
Raise your hand if you’ve undertaken a technology investment in the past several years that failed to realize the intended business outcome. Was there strong alignment from the beginning on the outcome, and was it shared across teams?
CDPs alone aren't driving personalized marketing experiences
According to Gartner’s survey data over the past two years, 42% of CMOs believe their organizations deliver 1:1 personalized messages to customers, 14% of organizations believe they have the right data to achieve a Customer 360, and about 1/3 believe they have the resources to execute their goals. Budgets declined 3.5x faster this year than they did last, reaching a three-year low.
Faced with a diminishing budget, limited and insufficient resources, and sustained or heightened performance expectations, the worst thing a marketing leader could underwrite is an investment without a crystal clear business case.
And yet, our collective industry — and the vultures that circle enterprises’ seven-figure purchasing decisions — have done marketing leaders no favors in helping to either disambiguate their options or deliver business outcomes, precisely because they are compensated to ambiguate options and are not compensated to deliver business outcomes.
As an example, here is a word cloud of three industry-leading CDP thought leaders’ buyer’s guides:
I was not surprised to see the prioritization of technical jargon (albeit I expected “composable” would be more prevalent).
I was surprised to see precisely zero mention of the following:
- Revenue
- Acquisition
- Retention
- “Use case” (insane)
- ROI / ROAS
The list goes on.
Another set of related trends has emerged alongside CDPs shifting focus to primarily service data and IT stakeholders:
- Organizations are using fewer CDPs
- Yet, organizations, on average, still use multiple CDPs
According to the CDP Institute, the average number of CDPs per organization dropped from 2.9 to 2.1 in 2024.
While the use of multiple CDPs historically involved a problem of category definition, I’ll explain here why this is shifting, how it relates to the above trend, and why it matters.
Organizations “using multiple CDPs” usually meant that the enterprise utilized multiple tools on the spectrum of data collection, organization, and activation. Category bundling — of tag managers, rETL tools, identity resolution-focused tools, and, in some cases, analytics tools and even marketing hubs as CDPs — explained this stat.
Now, organizations are consolidating tooling across the board and the workflow of their CDP into their marketing hub. I wrote about this when Klaviyo first announced their “CDP” and, more recently when Braze announced their data platform.
The real elephants in the room, however, aren’t the multi-channel marketing hub solutions whose ICP ranges from Shopify stores to high-scale digital natives but rather the enterprise data clouds and how these solutions are positioned for the next wave of AI-enabled marketing.
As I wrote after Salesforce launched its data cloud, the company's strategy is to use Matthew McConaughey to convince you that you still own your data (and you do) when you store it in Salesforce Data Cloud and then lease back access to your data for anything you want to do with it.
Here’s the thing: business stakeholders don’t care where the data is stored, if the application is composable, etc. They just want access to it to unblock their use cases and to drive business value. For this reason, we’re seeing a combination of:
- IT-focused CDPs and rETL solutions used by data teams to organize data within a data warehouse, and
- Business-user-focused CDPs used to power marketing use cases and the analytics around them (that copy and store data within their own ecosystem and charge you rent)
While this combination of solutions effectively solves the needs of each constituent group, the decision to use an “Enterprise Data Cloud” solution in addition to centralizing data within a cloud data warehouse is suboptimal from a cost and architecture perspective.
This leaves business stakeholders abandoned by the category and forced to choose between only bad options.
While modern CDPs may have lost the script in delivering business value, some, like Simon Data, remain substantially focused on the strategy, operations, and technology requirements of providing an effective customer marketing program.
We at Simon have not always gotten this right, either. I’m not yet ready to make New Year's resolutions for 2025, and I have too many personal flaws to prioritize CDP positioning even if I were, but as a business, we are investing significant support in this area for our customers.
The CDP was always intended to be a solution for marketers to access data and unblock data-driven use cases that drive measurable marketing outcomes. However, along the way, and with good reason, the focus has shifted toward this job's technical requirements and capabilities.
The next frontier of AI-driven and enabled marketing will require not only adherence to the technical underpinnings (which I do not mean to diminish) but also relentless focus on the business outcomes such initiatives are intended to achieve.
