Snowflake recently published its 5th annual Modern Marketing Data Stack report.
The latest edition’s theme is “governing the agentic enterprise,” and finds that AI is no longer just assisting marketers, it’s beginning to act independently. The key question, according to the new report, is whether organisations have the data infrastructure and governance frameworks to make that trustworthy at scale.
New Digital Age spoke with Erin Foxworthy, Global Head of Marketers & Agencies, Snowflake to find out more…
Why did Snowflake commission/conduct this research? What were the motivations?
The Modern Marketing Data Stack report is an effort that Snowflake started in 2022, and we just published our 5th edition. It has always been about helping marketers navigate major shifts in technology, and this fifth edition reflects the biggest shift we’ve seen yet: the move from AI that assists people to AI that can increasingly take action on their behalf.
What keeps us publishing is that marketers tell us they use it to benchmark and sharpen their own data stack strategy.
Marketers are under pressure to do more with their data while also meeting rising expectations around privacy, governance and measurable business outcomes. We wanted to explain how leading organizations are adapting their marketing technology stacks to support this new reality.
One of the motivations and conversations we wanted to have with this report is how AI isn’t primarily about adding another tool. It’s all about building the right foundation. Organizations that invest in governed, high-quality data and architectures that allow AI to operate responsibly will be in a much stronger position than those that simply layer AI onto fragmented systems.
What, for you, were the standout findings from the research?
One thing that stood out to me is that the most sophisticated marketing stacks are now organized around data instead of applications. AI is accelerating that shift because intelligent systems are only as effective as the data and governance behind them.
We also saw that governance has become a growth enabler rather than simply a compliance exercise. As organizations begin adopting agentic AI, governance isn’t slowing innovation down. It’s what makes trusted automation possible at scale. It matters on two levels: it keeps automated output accurate and trustworthy enough to act on, and it protects the proprietary data and brand IP that set a company apart, which only grows more important as AI flattens differentiation across the industry.
Another important finding is that marketers are moving beyond experimentation. The conversation has shifted from, “Should we use AI?” to “How do we operationalize it responsibly and demonstrate ROI?” That’s a much more mature discussion than we saw even a year ago, and it signals that AI is becoming part of day-to-day marketing operations rather than a standalone innovation project.
Did any of the research findings surprise or shock you in any way?
What surprised me most is that the biggest barrier isn’t the AI models, it’s organizational readiness.
Several experts we spoke with highlighted that unclear ownership, inconsistent data definitions and undocumented business logic are often bigger barriers than the technology. In many cases, companies have plenty of data, but not enough shared context for AI systems to make reliable decisions.
Another encouraging insight was that organizations don’t need to wait until everything is perfect before realizing value. Many are operating in hybrid environments, combining legacy systems with newer AI capabilities, and still making meaningful progress by strengthening governance and focusing on trusted data foundations first. For example, rather than ripping out a legacy data warehouse, a company might leave it in place and layer new AI capabilities on top, unifying and governing its customer data first so those tools have a trusted foundation to draw on. That sequencing lets them show value early without a disruptive overhaul.
What are marketers’ biggest misconceptions about leveraging AI and agentic tech (and the role of data alongside those)?
A common misconception is that AI alone creates competitive advantage. In reality, your data is your advantage, and AI amplifies whatever foundation already exists. If your data is fragmented, poorly governed or inconsistent, AI will amplify those problems, not solve them.
Another misconception is that agentic AI is primarily about automation. While automation is important, the bigger opportunity is enabling better decision-making at scale. That requires AI systems to operate with the right context, permissions and governance, not just the ability to complete tasks autonomously.
Ultimately, data isn’t just an input into AI. It’s the foundation that determines whether AI can be trusted. The organizations seeing the greatest success are treating governance, data quality and interoperability as strategic investments that enable AI to deliver measurable business outcomes.
Beyond AI, are there any other trends bubbling up that may be worth paying attention to?
Several trends stand out. First, we’re seeing the continued shift toward composable, cloud-native architectures. Organizations want technology ecosystems that are flexible, open and centered on shared data rather than isolated applications. Minimizing how often data gets copied is becoming a core principle in how teams design their architecture. Also, applications should return data back to the marketer’s own foundation to continuously strengthen it. When interaction and outcome data flows back into your governed data layer, it will fuel ongoing learning and optimization over time instead of value staying siloed in point solutions.
Second, privacy is becoming an operational capability rather than simply a regulatory requirement. As AI adoption grows, organizations are embedding consent, identity and governance directly into their data infrastructure instead of treating compliance as a separate workflow.
Finally, collaboration is evolving in important ways. Technologies like data clean rooms and secure data collaboration are enabling brands, publishers and partners to work together while maintaining governance and protecting sensitive information. As marketing becomes increasingly data-driven, the ability to collaborate securely across organizations will become an even more important competitive advantage.
