For most of the last fifteen years, the chief data officer role has been defined by a construction project. Consolidate the sources, stand up the warehouse, migrate to the cloud, and the value follows. The numbers suggest otherwise. In Deloitte's 2026 Chief Data and Analytics Officer survey of 100 data leaders, nearly all CDAOs, 95%, said their organization is not fully leveraging the value of its data, while 56% reported "intense" pressure to prove the business value and direct ROI of data and AI initiatives.¹ Put simply, fifteen years of platform building has not closed the gap between what companies own and what they use.
On a recent episode of AI Radicals, Alation sales leader Susan Wilson sat down with Cara Tice, Chief Data Officer at Early Warning Services, the company behind Zelle. There, fraud intelligence models run on consumer payment data inside a payments network that moved more than $1.2 trillion in 2025.² Tice has built the data function from nothing at three financial institutions before this one, which gives her a useful vantage point on data governance in banking. She has made most of these mistakes personally and is unusually willing to say so.
Why the one data platform strategy keeps failing data leaders
Tice's position on platform architecture is blunt. "I think that the monolith is kinda gone away, and now it's about the Lego blocks," she told Wilson. "But you shouldn't have too many Lego blocks."
Best-of-breed selection is now the default, and the failure mode has moved with it. Nobody regrets choosing the better tool. They regret the number of tools and the data integration work nobody scoped. "There's not one silver bullet," Tice said. "If I knew it, I would buy it, but there's no silver bullet. Data has many various capability needs, and so picking the best-of-breed solution is always a smart thing. But don't forget integration."
What matters more than the tooling, she argued, is the change process around it. "The most important thing that you can invest in is probably the change management program that's gonna surround it," she said. She then named who actually has to change. Data science modelers are moving to a new platform. Business users are building the data literacy to understand what data means through a catalog. Stewards are working a queue of incoming data quality issues. Across every build she has led, that program determined the outcome. "Every single time what gets us is the change management program, getting everybody on board, understanding how to use the tools. Getting value out of it. Measuring the value out of it."
Gartner's research points the same direction. In a March 2026 survey of 223 data and analytics leaders, Gartner found that cultural resistance outweighs funding constraints as the primary reason governance initiatives fail, 60% versus 40%. Gartner also predicts that by 2027, 60% of organizations that fail to address those cultural challenges will fail to govern AI successfully.³ Two organizations can buy identical tooling and get opposite results, and the difference is usually data culture.
Data governance hasn't changed, but its purpose has
The governance foundations most organizations are building on are thinner than the AI roadmaps stacked on top of them. Gartner found that 63% of organizations either do not have or are unsure whether they have the right data management practices to support AI. It predicted in early 2025 that through 2026, organizations would abandon 60% of AI projects unsupported by AI-ready data.⁴ The common read on that gap is that data governance needs reinventing for a new class of consumer. Tice rejects the premise.
"AI's been around for a long time, we've just called it different things," she said. "Predictive analytics, business intelligence, machine learning, all of that is on the spectrum of AI. And so data governance hasn't really changed in all this time. We want data owned, we want a common language that this term equals that term, we want to have good quality controls to make sure that the data is accurate through the system."
Governed data as enablement is the shift she describes. The requirements are the same ones practitioners have listed for a decade. The objective has moved from controlling access to data enablement, publishing data for use whether the consumer is an analyst, a model, or an agent. Gartner now gives data and analytics leaders the same advice, recommending they rebrand data governance as a business enabler and a team sport.³ As Tice put it, "AI eats data for breakfast, lunch, and dinner."
Governance work also needs a visible destination. Tice argues that data strategy has to connect to the company's OKRs, and she is direct about what happens without that link: "It feels like a librarian exercise to catalog CDEs. And while that is important, and I'm speaking to all my data governance friends, very important, you need to connect it to the why." Managing critical data elements in financial services only earns its budget when it's tied to an outcome. She makes that connection with a grocery store analogy she uses with non-technical stakeholders. The elements in your data catalog are ingredients, and the catalog is the labeled grocery store. The alternative is an analyst reaching into the back of the pantry for something they think is the right thing.
What a data product actually is: tables, joins, and an owner
Most definitions of a data product arrive as an architecture diagram. Tice's is deliberately unglamorous.
"A data product is all about tables and joins with context around it," she said. "You have an owner, you have quality rules embedded."
Data products work, in her account, because they turn governance from a process people comply with into an asset people consume. Ownership is the part most teams skip, which is why a data product operating model matters as much as the products themselves. "It's always been the strategy that everybody could understand that you're embedding governance into the data. You're eating the vegetables without knowing it."
They also address a quieter failure. Tice named one of the most frequent causes of data quality incidents, and it is not the data. "Oftentimes the issue around data quality is 'cause I pulled it from the wrong place. And like I said, this is everywhere." She says it's a pattern she sees at many companies, not only Early Warning. Data lineage makes the right source visible, but Tice's fix goes further. She builds internal-facing data products for consistent reuse, so models and customer-facing products draw on the same governed source. That's the same logic behind publishing to a data products marketplace rather than leaving consumers to discover tables on their own. For teams working out where to start, sequencing matters more than tooling, which is the argument in this guide to building data products.
Why data stewardship is the most under-resourced role in governance
Tice calls herself America's data steward, and she draws a line between stewardship and governance that most frameworks blur. Governance is the policy, standards, and capability layer. Stewardship is the hands-on custody of data as it moves.
"Stewardship is really about making sure data is quality controlled," she explained. "We have data owners, we have data custodians. If you're building your data governance framework, you have all these different personas. The steward, which is something you can really easy understand, they're about the care and feeding of that data through the entire ecosystem."
She has placed stewards at the system level, the process level, and the domain level across different organizations. Her conclusion is that the position on the org chart matters less than the existence of a named person. The role of the data steward is where governance becomes operational, because someone has to work the queue that automated data quality monitoring generates.
She is also realistic about the standard those stewards work toward. "Data's not perfect, and if you think it is, I could show you a thousand reasons why it's not." Her rule for how much validation is enough splits by audience: "progress over perfection always, but when it comes to customer-facing deliverables, we do strive for perfection." Either way, the output drives a decision. "Whether your consumer is a data scientist or is a commercial consumer, that data is used to make a decision, so you need to make sure that it's accurate." Internal work ships iteratively. Fraud intelligence delivered to financial institutions gets the full accuracy bar.
How to fund a data strategy in vertical slices
Every new data leader inherits a history of attempts. Tice's first move is always the same: a refreshed enterprise data strategy.
"If you're in a CDO role and you don't think you need an enterprise data strategy, then maybe the CDO is not a role for you, because it is so fundamental."
What she refuses to inherit is the instinct to rebuild. "What I would leave, let live and leave, is maybe some of the data assets that just need some improvements, not an overhaul. Don't jump to the conclusion that you need a one data platform, right? You might have it, and don't discount the prior eras before you until you take a look at it."
Vertical slices are her answer to the funding problem. The single-platform approach front-loads cost and back-loads proof. That produces the conversation she described: the CFO asks where the value is, and the honest answer is that another program is needed to get data back out of the box. "So many CDOs have the trauma story of putting all the data in one box," she said. Delivering value frequently in small chunks is, in her words, "honestly the thing that has made me more successful than the long tail one data platform to solve them all." Against the 56% of CDAOs under intense pressure to prove ROI,¹ that cadence is also the most reliable way to build a credible case for data governance ROI before the budget cycle closes.
Her advice for a first year is correspondingly narrow: "Fix a problem that has been plaguing the business. It's usually a data quality issue. Fix one. Like, fix it. Fix it, and fix it forever." The survey data backs that priority. In Deloitte's study, 61% of CDAOs said improving data quality and access was key to the success of AI and agentic AI initiatives.¹ The window Tice gives is three to six months to nail the solution, even if implementation runs longer.
Why AI governance depends on your data management program
Gartner expects spending on dedicated AI governance platforms to reach $492 million in 2026 and pass $1 billion by 2030, and much of that budget is being built as a category of its own.⁵ Tice treats AI governance as the same program at higher stakes. "I think it's an extension of data governance, personally," she told Wilson. Asked for the one thing listeners should remember, she was specific: "Your AI governance is reliant on your data management program. So make sure you've got data governance up and running, you're in the process to mature it, because AI is gonna rely on your data management controls and the context you create."
She also reframes what a governance program is for in an agentic context. Rather than a gate on access, it becomes the forum where use cases get argued: a "clear forum to debate the AI use cases" and confirm they are an appropriate use of the data. Her reasoning is that the failure is not recoverable. "You could inadvertently expose data in the wrong way, and once it's gone, it's gone."
For anyone about to scale agents, her most operationally useful advice reaches back to robotic process automation, where organizations deployed bots and lost track of them:
"I think it's really important that if you're embarking on AI governance that you have a strong center of excellence around AI agents and making sure that you have an inventory, you categorize them, you know what they're doing, that they can do one thing, can't do another."
Gartner makes the same point about governance tooling, calling a centralized AI inventory foundational because it lets organizations track every AI asset across its lifecycle.⁵ An inventory, a categorization, and an explicit scope per agent: that short list separates an agent estate from agent sprawl. It's what agentic data governance and an AI operating system approach are designed to make tractable.
What this means for data and AI leaders
Tice's argument is that the CDO has been promoted whether or not the role description caught up. "You cannot just stick in your lane and keep your head down and build that one data platform," she said. "You are a business leader, and you are enabling your company's customer outcomes."
That changes what good looks like. Start from the customer outcome rather than the source system inventory. Deliver proof in slices the business can feel. Treat the AI program as the existing data program held to a higher standard. Her advice on pace is the same one she says she has given at multiple industry forums: start small, start safe, get wins, test and learn. For leaders who want a concrete read on whether their foundation supports that, this data and AI readiness checklist is a reasonable place to begin.
Curious how governed, AI-ready data products can support your strategy? Book a demo with us today.
Sources and notes
Every external claim on this page is independently verifiable. The public sources are listed here.
In a survey of 100 C-suite data leaders at companies with $1B+ in revenue, 95% said their organization isn't fully leveraging the value of its data, and 56% feel "intense" pressure to prove the ROI of data and AI initiatives. 61% said improving data quality and access was key for AI and agentic AI success. — Deloitte, 2026 Chief Data and Analytics Officer Survey, 3 March 2026 ↗ https://www.deloitte.com/us/en/about/press-room/chief-data-and-analytics-officer-survey-finds-cdaos-acting-as-ai-trailblazers.html
Consumers and small businesses sent more than $1.2 trillion on Zelle in 2025, up 20% year over year. Zelle is owned and operated by Early Warning Services, LLC. — Zelle / Early Warning Services, 11 February 2026 ↗ https://www.zelle.com/press-releases/zelle-posts-20-growth-12-trillion-sent-far-outpacing-consumer-spending-and-cementing
A March 2026 Gartner survey of 223 D&A leaders found that cultural resistance outweighs funding constraints as the primary reason governance initiatives fail (60% vs. 40%). Gartner predicts that by 2027, 60% of organizations that fail to address governance culture challenges will fail to govern AI successfully, and recommends rebranding data governance as a business enabler and a team sport. — Gartner, 21 September 2026 ↗ https://www.gartner.com/en/newsroom/press-releases/2026-09-21-gartner-predicts-60-percent-of-organizations-that-ignore-data-governance-culture-challenges-will-fail-to-govern-ai-successfully-by-2027
63% of organizations either do not have or are unsure if they have the right data management practices for AI. Gartner predicted that through 2026, organizations will abandon 60% of AI projects unsupported by AI-ready data. — Gartner, 26 February 2025 ↗ https://www.gartner.com/en/newsroom/press-releases/2025-02-26-lack-of-ai-ready-data-puts-ai-projects-at-risk
Spending on AI governance platforms is expected to reach $492 million in 2026 and surpass $1 billion by 2030. Gartner identifies a centralized AI inventory as foundational to AI governance platforms. — Gartner, Q&A with Lauren Kornutick, 17 February 2026 ↗ https://www.gartner.com/en/newsroom/press-releases/2026-02-17-gartner-global-ai-regulations-fuel-billion-dollar-market-for-ai-governance-platforms
Analyst attributions and disclaimers
Gartner Press Release, "Gartner Predicts 60% of Organizations That Ignore Data Governance Culture Challenges Will Fail to Govern AI Successfully by 2027," 21 September 2026.
Gartner Press Release (Q&A with Lauren Kornutick), "Global AI Regulations Fuel Billion-Dollar Market for AI Governance Platforms," 17 February 2026.
Gartner Press Release (Q&A with Roxane Edjlali), "Lack of AI-Ready Data Puts AI Projects at Risk," 26 February 2025.
Gartner does not endorse any vendor, product or service depicted in its research publications and does not advise technology users to select only those vendors with the highest ratings or other designation. Gartner research publications consist of the opinions of Gartner's research organization and should not be construed as statements of fact. Gartner disclaims all warranties, expressed or implied, with respect to this research, including any warranties of merchantability or fitness for a particular purpose. GARTNER and Magic Quadrant are registered trademarks and service marks of Gartner, Inc. and/or its affiliates in the U.S. and internationally and are used herein with permission. All rights reserved.
- Data Catalog
- Data Governance
- Data Products
- Data Intelligence
- Digital Transformation
Keep reading
More from the data desk
