Data is growing at an exponential rate, with Deloitte predicting it will reach 175 zetabytes by 2025—the equivalent of one billion one-terabyte hard drives.

Our expanding data stores are helping create new opportunities for innovation—through AI, machine learning, IoT, and generative AI. To get more value out of this data, organizations need an end-to-end data strategy that encompasses more than just technology. Successful data strategies include upskilling workers, creating a data-driven culture, and fostering data-driven decision-making and innovation.

More than 80 percent of large companies today have incorporated a Chief Data Officer (CDO) role into their C-suites to lead these elements of data strategy. Ideally, the CDO position oversees a range of data-related functions to ensure an organization is getting the most value from its data. With the advent of generative AI, companies are also looking to their CDOs to understand this new technology and develop strategies for using it productively and responsibly.

In a recent study commissioned by AWS, CDO Agenda 2024: Navigating Data and Generative AI Frontiers, researchers and thought leaders in the data and AI space set out to understand how the CDO role is evolving—and what successful data leaders do differently from the rest. Five key themes emerged.

Business outcomes come first

Forty-four percent of CDOs say they define success as achieving business objectives, as opposed to technical accomplishments, and mark analytics and AI as keys to providing value. Business objectives can be customer experience improvement, risk reduction, employee motivation, or specific ROI figures. Whatever objective is chosen, CDOs put KPIs in place to track their progress on that goal and then let those KPI scorecards provide visibility into results that inform future initiative spending.

“When I first joined Vista,” said Sebastian Klapdor, EVP and CDO of print and graphic design company Vista, “there was a central team of engineers sitting in a physical data warehouse in a basement, and that just does not scale. The first thing we did was adopt a data product approach—treating data like a product and strategically developing, launching, supporting, and ensuring success of data products within the organization—and federated it across data domains along our customer journey. For each domain, we created a value pool—for marketing, pricing analytics, manufacturing, customer care, and so on. Depending on these value pools, we started to think about the biggest and most valuable problems we could solve with data, with analytics, and with AI. Then we took the first problem from each list and ramped up data products that could scale and deliver value.”

A data-driven culture is crucial

Before an organization can use data, its people need to understand data and how to apply it to the business. Creating a data-driven culture is a major focus for over half of the CDOs surveyed, beginning with data literacy programs and change management. And CDOs agreed that an organization’s leaders must set the bar.

"It really helps when you have a four-star general and former chief of space operations talking about how he's taking a Python class," noted Eileen Vidrine, the Chief Data and AI Officer for the United States Air Force. "That's leadership by example.”

Building a data-driven culture also means creating training programs that meet your workforce where it is—and sometimes your workforce has skills all over the map. At the Air Force, for example, the workforce represents a diverse range of ages and educational backgrounds. Vidrine wants to ensure that every member of the team has opportunities to learn no matter what their current skill set is.

Klapdor further highlighted the idea that data-driven transformation isn’t a solo enterprise. “You have to have data literacy outside of the data professionals,” he said, “because you don't only want the data people to know data—you want everyone who needs to drive a decision to have the right data and know how to use it.”

Having identified “being data driven” as a core value for the company, Vista built data-driven evalsuation into its interview guidelines and performance evalsuations. "Every one of the data professionals we have at the company serves as a data ambassador," said Klapdor. "They help their colleagues be more data literate and understand the value of data.”

An enablement approach governance

CDOs spent 63 percent of their time in 2023 focused on data governance, up from 44 percent in 2022. When it’s done right, data governance empowers people across an organization to innovate and derive insights from internal data. “We need to get the data into the hands of the people who use it,” Klapdor said, which means making data visible, accessible, and easy to use.

Organizations that make data-driven decisions operate more efficiently when data is visible and accessible, as opposed to being guarded by departments or gatekeepers. At the Air Force, Vidrine has empowered every airman to be part of the solution. "We had volumes of data that could drive insights if it was just made visible and accessible for our workforce," she said. She got serious about training and data literacy and then put systems in place to enable that access.

Be prepared for AI in all of its forms

Business leaders across all industries are excited about the possibilities and opportunities that come with the use of generative AI, and 93 percent of CDOs agree that having a data strategy is crucial for realizing the full potential of this new technology. Surveyed CDOs suggested starting out with generative AI by targeting a use case—large or small—and demonstrating tangible impact.

Although generative AI is the new buzzword in boardrooms, CDOs suggest starting with a customer problem that makes sense to solve with generative AI. “I start by talking to the leaders of different functions and saying, ‘Tell me your problems,’” explained Catherine Miller, CTO of New York-based healthcare technology company Flatiron Health. “I say, ‘Let's talk about the best tools for solving those problems.’”

At Vista, Klapdor’s team has already pinpointed areas where generative AI will be useful, from creating automated designs for customers to improving customer service help lines. “What we need,” he said, “is to enable our teams across the organization to experiment and to think hard about the different ways they can use this technology—and then make it easy for them to adopt and scale it in a safe and responsible manner.”

CDOs across the board see the need for building responsible AI and for having a sense of responsibility permeate the whole enterprise. At the Air Force, it’s the job of the responsible AI lead to make sure that “every single airman, every single guardian, our whole workforce, is thinking about building responsibly," said Vidrine.

Innovation for everyone

Successful CDOs have seen a tangible ROI on their data-driven initiatives by using a “think big, start small, scale fast” approach. From small, experimental starts, they add additional capabilities based upon need, and then scale fast and use data to inform their next steps—even in the face of scrutiny.

At Flatiron Health, this approach has been crucial. "We started small by finding an early partner that was willing to work with us on those initial, not particularly lovable data products,” Miller said. “That allowed us to go through the whole process of mapping and cleaning data and thinking about what questions we could answer—and how, over time, we could scale.”

For Klapdor, Miller, Vidrine, and the rest of the 350+ CDOs surveyed in CDO Agenda 2024, the ability to solve measurable business problems with data, smartly explore new technological frontiers, and create data-driven organizations have been crucial to their success. But they haven’t done it alone. “I like to say data and AI are team sports,” Vidrine said. “If you try to do it by yourself, you're not going to be as successful as you can. I think that whole collaborative team approach is the key to long term success.” Learn more about how AWS can help your business grow at AWS for Data.

This story was produced by WIRED Brand Lab for AWS.