Putting the ‘A’ in CDO: The rise of the chief data and analytics officer

In the data industry, there’s a new sheriff in town: the chief data and analytics officer (CDAO).

More than a job title, it’s a new perspective on how firms view data and its value—both inherent and derivative—as an offensive asset rather than a strategic or merely defensive tool. Additionally, its responsibilities reflect that data professionals are no longer just in charge of data management and governance, but are now also tasked with creating new tools that deliver insights based on the data they manage.

In the past, managing a firm’s data fell under the overall responsibility of the chief data officer, with market data and reference data staff running regions or business lines. So, should this new role eliminate CDOs? Or will their jobs evolve to incorporate new areas of responsibility to meet the new demands of changing times?

Asset manager Vanguard does not have a chief data officer, but it does have Ryan Swann, who has held the position of CDAO since 2021 when the firm created the chief data analytics office. According to Swann, the office’s primary responsibility is managing Vanguard’s entire data lifecycle—specifically, how the firm manages, stores, and uses data to give investors the best chance of success. 

Every day, we focus on how we incorporate responsible practices—think privacy by design—into how we use data to ultimately help our clients

Ryan Swann, Vanguard

Secondly, the office is responsible for descriptive analytics, predictive analytics, and advanced analytics, which include artificial intelligence, machine learning, data science, and behavioral science. Finally, the office also has responsibility for data governance and what the firm calls “data defense”, which ensures it is using data and developing analytics in a “responsible” way.

“Every day, we focus on how we incorporate responsible practices—think privacy by design—into how we use data to ultimately help our clients,” he says.

His role takes the traditional governance and data oversight responsibilities of a CDO, and applies them to new opportunities, taking an offensive strategy compared to the mostly defensive focus that CDOs have pursued in the past. So, where does a CDO’s role end and a CDAO’s begin?

Gary Kotovets, chief data and analytics officer at business data provider Dun & Bradstreet, sees his role as a combination of CDO with analytics oversight. Sounds straightforward, but it’s more nuanced than that.

“As a CDO, your typical role is to manage the back-end of an organization—getting data in, managing it, cleaning it, and serving it up to wherever the customer is. … And with my analytics hat on, I’m now also a consumer of that data,” Kotovets says. “I now see the back end and the front end at the same time, and I know the outcomes I want to produce when I look at the raw, incoming data.”

Thinking about data from a user’s perspective rather than from the perspective of a data manager is how a CDAO works differently. As a CDO concerned with governance and corralling data, you’re putting fences around it to keep it well managed, whereas a user wants to saddle up that data and see how far it can ride.

“I have to constantly think about how to help customers be aware of new risks that emerge in the world: supply chain risk, geopolitical risk, and others. So, when I think, ‘What data do I need to collect to mash up?’ it’s a much faster process for someone who wears both hats,” Kotovets says.

It also allows the CDAO to be more critical about ingesting data. Whereas a CDO’s concern may have been to get all the data right, a CDAO can look at what’s actually being used—or not—by internal or external consumers, trace it back to the ingestion point, and decide whether a dataset is being onboarded unnecessarily, and whether to discard it so as to not spend resources managing unnecessary data.

Limitless ways to unlock data

When Kotovets arrived at D&B in January 2020 after 19 years at Bloomberg overseeing everything from data management to the vendor’s EMS, just as at Vanguard, creating the CDAO role required some restructuring. At the time, D&B had just been taken private by a private equity firm, only to go public again just over a year later. Amid that and management changes, Kotavets set about building his ideal data organization.

“When I arrived, they deconstructed the whole data organization, so I had to start from scratch. They already had a chief analytics officer, so as I rebuilt the data organization, it was clear that analytics was very important,” he says. Ultimately, he combined the roles into the CDAO position because it proved more efficient when combining data and analytics into sellable products with shortened time-to-value.

I have to constantly think about how to help customers be aware of new risks that emerge in the world: supply chain risk, geopolitical risk, and others. So, when I think, ‘What data do I need to collect to mash up?’ it’s a much faster process for someone who wears both hats

Gary Kotovets, Dun & Bradstreet

“One of my first objectives was improving data quality, and understanding usage and client feedback—what are our clients communicating to us about the data? That allowed me to improve quality and reduce noise and unnecessary content that we were spending a lot of time and money acquiring and managing, but which wasn’t getting used,” he says.

Now, aside from its DUNS identifier, D&B provides data on more than 560 million businesses, adding around 4,000 data attributes per business. It provides details such as corporate hierarchy and structure; ownership; news; dozens of scores, such as payment performance, delinquency, and probability of a business failing. Together, this builds an understanding of a company’s performance and exposure to all manner of risks, including climate risk.

“We have so much data, and there are limitless ways to unlock it,” Kotovets says. Spotting those ways and making them happen is where a CDAO comes in.

Elizabeth Pritchard, CDAO at Liberty Mutual Investments, the asset management arm of US insurance company Liberty Mutual, says that even after existing for a couple of decades, she still sees the CDO role itself as a fairly new phenomenon, and a CDAO role—or other iterations of data, analytics, and AI—are even newer. She joined the company in 2022, after almost 20 years at Goldman Sachs serving in various capacities, including global head of market data services. She also spent two years as COO of AIG’s data science group, and was co-founder and head of go-to-market at Crux Informatics.

“Here, we had a part-time CDO in the technology organization, and I was hired as the first CDAO role within the business to build the capabilities for driving value within the business, working with technology,” Pritchard says. “The CDO role here—and, I think, in the industry generally—was created to get the right governance and access controls in place, and to stand up the right technology, storage, and pipelines for data … and that was a great foundation for what I’m trying to do as CDAO. The pipes and groundwork were already laid, and now I can drive value from data.”

Pipes and groundwork for what, exactly? What are firms building on top of this foundation of good data management and governance?

“The CDAO is more focused on driving stakeholder value. Why? Because we have the analytics,” Pritchard says. “Just having the raw data doesn’t make it easy for a business user or an investor to make decisions. They need analytics. To drive value, the data needs to meet the business user in a way they understand.”

Put another way, Pritchard sees her role as enabling business strategies. Whatever the strategy is, her job is to understand the implications of that strategy on the firm’s data and technology resources, and to use those to deliver appropriate analytical tools to end users, faster and more focused than before, to rival the quant and strategy groups assembled by large banks and hedge funds.

Just having the raw data doesn’t make it easy for a business user or an investor to make decisions. They need analytics. To drive value, the data needs to meet the business user in a way they understand

Elizabeth Pritchard, Liberty Mutual Investments

This means creating analytics platforms where users can run ad-hoc queries, rather than waiting in a queue for a firm’s technology organization to roll out a specific tool or dashboard.

“We and many others are moving fast down that path, so the analytics need to be responsive to ad-hoc queries. The big banks have had that capability at the core of their quant groups,” Pritchard says.

But even then, it may take 18 months to show value from a CDAO’s efforts, she says, depending on a range of factors, including the maturity and complexity of an organization and its investments, how data was organized previously, and the talent already in place. To get there quicker, she recommends starting with use cases specific to stakeholders, where incremental wins will demonstrate value, sooner.

Cementing responsibilities

The pipes-and-groundwork construction analogy is a good one: Aside from the similar timeframes involved in building a house and a CDAO function, some liken the distinctions to the different roles involved in building and selling a house: first, you need architects and construction professionals who design it and build the foundations and structure, install the plumbing and wiring, and ensure everything is up to code.

Then, only once that work is complete, do you employ someone else who is an expert in staging the house, making it attractive, valuing it, showing it to prospective buyers, and closing a sale. In this scenario, the CDO oversees the architecture, foundations, and structure, then the CDAO turns those assets into something marketable.

Over time—perhaps when an organizational structure changes and splits or augments roles—the CDAO may be required to redesign and rebuild those foundations to meet new requirements. Or, depending on the scale and maturity of a firm, and how far it has gotten in establishing a core foundational model and control functions with data and model stewards assigned throughout the organization, that function may become self-fulfilling and be shared across business areas, rather than centralized in one role.

But laying the groundwork for that decentralized approach often, ironically, begins with centralizing control under one individual precisely so that they can restructure and reassign responsibilities for the future. And change—even the initial creation of a CDO where one doesn’t already exist—doesn’t come easily.

One of the challenges of any centralized function, including the CDO role ... is that you start off with a lot of resistance, because with centralization, you’re taking away control from others

Evelyne Roy, Thomson Reuters

Evelyne Roy is head of data and analytics at Thomson Reuters, but before that, she worked as a consultant and held senior data roles at banks in Canada and Australia. In one of those prior roles, helping a bank implement a data governance function and structure for risk reporting, she advocated that those functions shouldn’t be run by the risk group itself. Rather, as an enterprise function, they should fall under the remit of the CDO.

“One of the challenges of any centralized function, including the CDO role ... is that you start off with a lot of resistance, because with centralization, you’re taking away control from others,” Roy says.

Appointing a CDAO is rarely a seamless transition. One bank where Roy previously worked first expanded the CDO’s remit when one individual left and another took over, then changed the job title completely to head of analytics when yet another person took over. There’s no one-size-fits-all approach, and what worked at one firm may not work at another.

And, just as with Pritchard’s assessment of the factors behind the success and timeframes of a CDAO’s function, Roy also says each situation is different and depends on the maturity and size of an organization, and its challenges and priorities. Therefore, she says, a CDAO can’t simply apply the same template approach across different firms.

Another area where Roy and Pritchard agree is that any CDAO-led effort must be distinct from, yet tightly knit with a firm’s CTO and its technology function.

“In banking, firms tend to segregate duties. And the platforms used by a CDO function may be run and supported by the CTO and technology organization, while at some firms, the CDO role has a technology responsibility for data platforms,” Roy says.

It’s a delicate balance, and the two leaders and their organizations need to work like dance partners—in lock-step and in harmony, with one side contributing the right data in the right format, and the other delivering tools that will display, deliver, and help users get the most out of that data.

“Historically, decisions may not have been so tightly coupled. Now, we may not need a CDO, but I will say that we do have a very strong, very tight partnership with the technology group,” Pritchard says. “So, the CTO is really important to enable that, and the CDAO is working with content and developing the data in such a way so that it’s available for those ad-hoc queries.”

The CDO role is still evolving. We refer to CDO 1.0, CDO 2.0, and CDO 3.0. I was one of the original CDOs, so I call myself ‘CDO beta’

John Bottega, EDM Council

Vanguard achieved that delicate balance by lifting and shifting individual data and analytics organizations—totaling more than 800 staff—into the CDAO office while remaining co-located within the business lines they’re aligned with to maintain close ties with the teams they support.

Vanguard’s Swann says this approach keeps business objectives aligned, breaks down silos, and establishes “a unified data ecosystem that enables seamless data integration, sharing, and collaboration across the organization.” In short, a centralized structure with distributed execution allows analytics teams to maintain their business-line focus while also working as a whole to deliver on enterprise-wide initiatives.

Old dogs, new tricks

John Bottega was one of the first CDOs on Wall Street, and certainly the most visible and vocal proponent of the role, which he held at Credit Suisse, Citigroup, the Federal Reserve Bank of New York, and Bank of America over the years—not to mention his prior experience of over 20 years in systems and data roles in finance before the CDO title emerged.

Bottega, now president of data industry body the EDM Council, says he doesn’t necessarily see the CDO and CDAO as separate roles, but rather that the CDAO represents an evolution of the CDO role to meet changing times and needs. “The CDO role is still evolving. We refer to CDO 1.0, CDO 2.0, and CDO 3.0. I was one of the original CDOs, so I call myself ‘CDO beta,’” Bottega says.

But market forces drove the role to evolve: first to a defensive position to tackle the financial crisis, then to more of an offensive role as the business side demands to know, “What are you doing for me now? What are you doing with all that data?” in terms of developing new products or improving customer service.

For example, at Vanguard, transforming data into a “strategic asset” enables more personalized client experiences, scales its ability to advise clients, optimizes investment and business operations, and reduces risk, Swann says. Its efforts to make data more accessible, actionable, and ultimately valuable enable the firm to equip all users with resources to make better decisions “regardless of their role, tenure, level, or expertise,” he adds. “Creating a culture in which employees are empowered to use data and analytics resources to help improve client outcomes and client experience will continue to be a focus area for us as we look to the future.”

It’s exactly these growing and more complex needs that have driven the emergence of the CDAO, Bottega adds.

“Data management is rapidly evolving, and so are the business drivers that impact it. So, it’s good that this role is constantly changing, morphing, and evolving. I think the CDO taking on responsibility for analytics is natural.”

Creating a culture in which employees are empowered to use data and analytics resources to help improve client outcomes and client experience will continue to be a focus area for us as we look to the future

Ryan Swann, Vanguard

As part of that “natural progression”—and associated title change—the CDO-turned-CDAO is now taking more of a role in innovation. After all, as the industry rushes to adopt new technologies such as machine learning and artificial intelligence, “there’s a recognition that these technologies are only as good as the data you feed into them,” Bottega says.

So, who better to ensure data is used properly when creating new tools and analytics than the same people who set up and defined what constitutes proper usage and governance in the first place? Just as with Vanguard’s definition of data defense, one must first understand the principles of data governance before they can apply it to the new types of use cases for data and analytics.

That said, elevating the role of the CDO into something new—while taking on overall responsibility and accountability for new areas—doesn’t mean that the CDO must do everything themselves. Like Roy and Pritchard, Bottega advocates for a structure where others in the organization have daily oversight of tasks like data governance, model governance, and outcome governance. In fact, one of the responsibilities of expanding the CDO role into a CDAO is bringing in others to execute those tasks, he says.

So, what’s next, and where does the role go from here? Duncan Cooper, CDO at Northern Trust Asset Servicing, says that while CDOs have typically been limited to large sell-side and tier-one buy-side firms, he expects to see CDOs become ubiquitous across the buy side over the next decade.

But, whatever their title, he expects the job descriptions to be much more of a CDAO-style role—or even CDDO (chief data and digital officer)—responsible not only for data quality, but also for packaging and getting that data to internal consumers, such as analytics that can help marketing and sales teams generate competitive advantage and drive decisions.

The CDDO role, he says, is another natural extension, taking the data and analytics created by the CDAO and incorporating the channels and user experiences by which those packages are delivered to customers.

“When you name a CDAO or CDDO, you’re adding commercial awareness, product development and management, marketing skills, and a grounding in the technology and technical capabilities. Being able to talk at the same level as technologists is a key skillset,” Cooper says. “Now we’re seeing that data can be an offensive asset—we’re entering a new generation and the CDAO is definitely the future role that we’ll see come to fruition.”