Dive into the changing dynamics of CDP ownership, usage and collaboration between marketing and other internal teams.
“Who should the primary user of my CDP be?”
“What should the team look like supporting my CDP effort?”
“Who should own this initiative?”
Brands adopting a CDP often ask me a version of these questions over my 10+ years in the space. I’ve trusted answers born of experience earlier in my career for a long while. However, a lot has changed in the marketing technology space.
It’s worth reconsidering who the user is for your CDP — especially in light of advances within “composable CDP.” Hint: It’s not one person or one role. More to come on that.
The rise — and evolution — of customer data platforms
Let’s recall what a customer data platform is. David Raab, the godfather of CDPs and the person credited with first recognizing its growing significance in marketing technology, deserves some recognition here. He founded the CDP Institute, which defines a CDP as “packaged software that creates a persistent, unified customer database that is accessible to other systems.”
Many CDPs have tried to embrace marketing as the core user. MarTech published this article, which calls out CDP as “marketer-managed.” This fits with my recollection of history.
2017 was a watershed year where I witnessed firsthand a massive influx of new clients to the Lytics CDP during my time there running customer success. At that time, it was accepted that CMOs would outspend CIOs on technology.
My how things have changed. Fast forward from 2017 to 2024:
- GDPR, CCPA and a rapidly growing patchwork of state-determined policies influence data residency, accessibility and proper usage. As such, customer data is firmly in the domain of the IT organization.
- Signal loss from third-party cookie deprecation is a growing threat to the marketing team’s ability to manage and measure their campaigns.
- New channels and formats to get a message across via influencer marketing and on video platforms like TikTok and connected TV now command more marketing attention.
- The Cloud Data Warehouses and accompanying data management tools like Snowflake and Databricks have spent considerable resources courting marketing use cases.
- Artificial intelligence advances a business users’ ability to scale segmentation and creative exponentially.
Marketers today are often confounded by these cataclysmic shifts in the landscape they operate. They need the owners of enterprise data to be a key partner in the journey. Thus, The CMO is a key influencer in the CDP decision that a committee — usually led by IT — makes. This is especially true in the composable paradigm, where IT-driven architectural paradigms govern marketing technology decisions.
There are exceptions to this rule — but let’s accept that the CDP decision is no longer a “marketer-managed” decision and accept the broad reality that CDP decision and usage have shifted as the buying dynamics have shifted over the last five years. Let’s also accept that the buyer and the user of the CDP may be different.
Who are the CDP users historically?
Historically, I used to contend that semi-technical users supporting marketing operations were the ideal users of most CDPs. These users were focused on supporting use cases, marketing data and campaign reporting. Importantly, they interfaced with enterprise data, analytics and marketers.
Yet, the CDP could not subsist on MOps alone, so this group required significant collaboration from engineers: data engineers to feed offline data to the CDP and front-end engineers to ensure key data, audiences and events were functional in marketing channels. Analysts and digital marketers rounded out the CDP center of excellence.
The former was on tap for complex reporting and insights, supplementary data enrichments and often for simple data feeds that were a lower priority for the enterprise data engineering team. The digital marketer usually made decisions about marketing use cases, such as how audiences and customer attributes would be used in ads, email, site personalization, push notifications, etc.
This paradigm mostly worked with the decision-oriented CDPs I’m most familiar with — Lytics, ActionIQ, BlueConic, Simon and Blueshift. But I noticed it breaking down with product-oriented clients where technical users were most commonly the CDP leads. We saw these more with clients on Adobe, mParticle and Segment — who have often courted CIOs and the technical organizations underneath them. With the more data and IT-focused platforms, we tended to see product engineers be the main users for data routing, with marketers as a secondary consideration.
What about composable CDPs?
There have been notable dialogue forced in the space by new entrants such as GrowthLoop, Hightouch and Census. Traditional CDPs such as Lytics and ActionIQ have responded by adding noteworthy reverse ETL capabilities in “Cloud Compute” and “HybridCompute,” respectively.
The composable paradigm is here to stay. Working with pre-existing data in a data warehouse offers theoretically faster time-to-value, lower total cost of ownership (TCO) and improved data security since key data does not need to be replicated in another system.
Composable CDPs encounter several hurdles in reaching enterprise value. Technology changes much faster than the rate of human beings’ ability to adapt to those changes. Political and organizational friction exists inside companies — even small ones.
Changing a process to incorporate first-party data better can be difficult — and a burden of proof of being the challenger to “the way we do things.” But there are some things that adopters of composable CDP ought to consider.
1. The product owner
Composable CDPs are the client’s product. It’s their database and the composable CDPs are working primarily to unlock key capabilities in the data warehouse to support first-party data-driven use cases, marketing analytics and new features in the database.
This person can highlight key missing data elements of the customer journey and organize internal workstreams to address gaps in capabilities.
2. The customer insights/marketing analytics team
As CDPs have grown in prevalence in the enterprise marketing stack, so too have the focus of enterprise customer insights teams on first-party data activations. During the early days of CDP, the customer insights team was often a dubious detractor who focused on bespoke insights from customer data that didn’t connect tightly to downstream, in-channel activations.
By 2024, enterprises have determined that the customer insights team has better command of the data. The customer insights team can usually better interpret data nuances to business value in a way that today’s marketers cannot. These teams now must support segmentation, real-time decisioning and journey orchestration in a manner not conceived in the “marketer-managed” paradigm of yesteryear.
A key component of this is AI-driven segmentation. The customer insights / marketing analytics team sits in a great place to vet how to best leverage internal data science and capabilities associated with marketing technology tools or the CDP itself.
ML engineering for generative AI
Composable CDPs, built directly on a client’s cloud data warehouse, offer a unique ability to blend generative AI capabilities of the cloud with marketing campaigns in a seamless way.
The barriers to creating good generative AI on top of the existing tools that OpenAI and Google Cloud provide are low. I can say, with firsthand experience, that one of my companies, Predictable, has built very good POCs of generative AI capabilities inside three sprints.
Composable CDP adopters should consider enabling business users to build segmentation using generative AI to democratize further segment-building to in-channel marketers.
AI-generated customer-facing creative for marketing campaigns will face a tougher challenge. In the near term, generative AI in customer-facing applications will be confined to subject line testing, basic copy, call-to-action tests and subtle personalization of images for highly segmented campaigns. However, much of this may fall to the channels themselves. For example, generative AI in Google Ads, your ESP for subject line testing or your Adobe Creative Cloud for image assets.
What aspects of the center of excellence are the same with composable and traditional CDPs?
Whether your strategy is to build a composable CDP of multiple complementary technologies or choose a CDP that provides all CDP capabilities in one package, it is a fundamentally cross-functional effort.
- Executive sponsor. The adopter of a CDP must have a clear top-down remit of achieving value and relevance from their first-party customer data as a key workstream in a digital transformation initiative.
- Data engineering. You will still need data engineering to be a key stakeholder to feed a traditional CDP or to engineer a customer data warehouse with minimum viable data or better.
- Front-end engineers still need to help digital marketers build more integrated personalization. This can occur with data feeds from on-site and in-app events or detailed email customization.
- Marketing operations will still be on hand to ensure smooth campaign execution.
- Legal / compliance partnership. Educate and bring your legal team along in the journey. There are ways to build compliance and consent into your process that build trust with this important constituent.
- Roadmap. Have a roadmap of use cases and data capabilities to add. Focus on low-effort, high-value wins early and build use cases and capabilities based on an objectively scored vetting of all feasible options.
- Governance. Build in a process to ensure that consent attributes are followed, audiences have good structure, campaigns are measurable, etc.
Be intentional about your center of excellence. Build a complementary team with diverse skillsets and make sure you stay current with this rapidly evolving landscape.
The post Building a center of excellence for CDPs in 2024 appeared first on MarTech.
MarTechAbout the author
Craig Schinn is Co-founder of Actable, a consultancy focused in helping clients organize, analyze, and activate 1st party data. Craig has over 20 years of professional experience in consulting, marketing analytics, and marketing technology. In 2014, He was an early adopter of CDP during his time at an online retailer, and has been invigorated by helping clients get more value from 1st party customer data.
(13)
