When Steve Stine became AT&T’s first chief data officer (CDO) in 2017, the company’s analytics operations were scattered among different teams in charge of data streams, analytics, and automation. Data systems weren’t widely accessible and were often difficult to use.
Stine brought together the teams, led efforts to make systems more user‑friendly, and introduced new analytical tools. With the tools, AT&T’s retail operations were able to identify new‑store sites more efficiently and increase revenue at existing locations. Meanwhile, AT&T’s field‑service operations team used machine learning to identify the best travel routes for roughly 43,000 AT&T technicians.
Executives like Stine—who was recently succeeded by new data chief Jason Porter—occupy one of the newer offices in the C‑suite, and increasingly one of the most important.
Chief data officers rose to prominence in the wake of the 2008 global financial meltdown, according to Deloitte, as regulators pressed banks to assign a high‑level executive to vouch for the quality of mortgage and asset data.
Today, as data accumulates in unimaginable volumes, organizations realize that the mountains of information aren’t just relevant to immediate business problems. Increasingly, big data is critical for charting long‑term success. Accordingly, companies need executives who can both safeguard that data and tap its full strategic value.
“There needs to be someone to unlock the value of that data and embrace data culture,” says Jojy Mathew, a principal in the Analytics & Cognitive practice at Deloitte. “The role of the CDO is increasingly about leveraging data as an asset.”
Today’s chief data officers face a tough juggling act. One part of their mission is simply keeping data secure, checking that it conforms with regulatory requirements and ensuring its integrity.
CDOs also need to be agile innovators, leading—and sometimes pushing—their organizations to put data at the center of all business processes. The rapid adoption of artificial intelligence to streamline those processes and drive significant future revenue is further elevating the CDO’s profile.
About 57% of U.S. companies reported having a formal chief data officer role in 2017, according to a recent Gartner survey. That’s a 50% jump from the prior year. A third of the CDOs surveyed said increasing revenue was their main measure of success. Moreover, companies that hire a CDO are more than twice as likely to have a clear digital strategy than companies that do not according to similar research from KPMG.
Given the relative novelty of the title, it’s not surprising that chief data officers face a number of challenges. In many organizations, the job is poorly defined, and responsibilities often blur with those of chief information officers and chief technical officers.
The CDO role usually requires the ability to drive organizational change, not just implement new technologies. Here’s a look at some of the key issues chief digital officers are dealing with and how they address them.
Creating a data-driven organization
To be successful, CDOs must push their organizations to place analytics and data‑based decision‑making at the center of everything they do. That requires leaders to both adopt technology and build cultures that enable everyone from the CEO to field workers to incorporate data into their daily work.
On the technology side, it means systems that make data more accessible and tools that put advanced analytics in the hands of people across the organization.
“The role of the chief data officer has evolved beyond gathering data,” says Kim Keating, AT&T’s vice president of data science, who reports to the chief data officer. “Collaboration has been central to that effort. By teaming up with organizations across our company, we’re able to use our data in smart and innovative ways that enable us to drive operational efficiencies and improve customer experience.”
Stine, now senior vice president for business transformation, also led efforts to apply artificial intelligence to the problem of identifying network problems—and anticipating them before they occur. The company receives 15 million outage alerts daily. With AI, it “can predict outages before customers are impacted and call in to let us know,” he wrote in a 2018 blog post.
As CDO, Stine also championed the reskilling of employees to help them use data more effectively. He led efforts to train workers in machine learning, advanced analytics, and bot building. Today the company has 2,000 employees working on automation projects, and has more than 1,000 bots deployed throughout its various operations. These bots help technicians install equipment for customers and aggregate data for service orders. One bot scans customer‑service phone calls and compiles reports on network traffic.
Making data work
Data management remains the first order of business for the chief data officer. In many cases, that means synthesizing data in many formats from hundreds of sources.
Given the vast volumes of data that governments collect on everything from housing and healthcare to education and transportation, it’s no wonder that state and local agencies have been at the forefront of hiring CDOs to make the most of all that information.
Maksim Pecherskiy is chief data officer for the city of San Diego. Pecherskiy says his first job is to identify value in the city’s myriad data streams. One product of his work is the city’s StreetsSD website, which displays real‑time road conditions and repair work.
Pecherskiy’s team created StreetsSD using data from the city’s Pavement Management System crews, geo spatial mapping data, and historical street data. They set up “Airflow,” an open source Apache program developed by Airbnb, to automatically combine data sets for use.
Not all the work could be automated. Technicians and city staffers, including personnel at the Transportation and Storm Water division, clean the data manually and upload it in specific formats to make the project work.
Pecherskiy and his team also draw on advanced analytics to improve ambulance and fire dispatch in the city. In the past, dispatchers often sent too many or too few ambulances or fire trucks in response to emergency calls. By standardizing data from multiple sources, the city significantly improved the efficiency of emergency dispatch.
“Now, someone who has a severe emergency call gets the correct level of resources,” Pecherskiy says. “And the city saves money by not sending too many resources to less critical emergencies.”
Looking around corners
In addition to using data to solve pressing business problems, CDOs must also use analytics to identify business opportunities. This shift is reflected in the Gartner survey, where more than a third of the respondents said increasing revenue is a main way they measure success. CDOs in the survey said they spend 45% of their time focused on generating revenue or creating business value from data initiatives.
Consider insurers. One Forrester report notes that in the 1970s, insurance companies were hit by a wave of unanticipated litigation over the health risks of asbestos. Today, tapping the power of AI and advanced analytics, a CDO could search through academic and medical studies, legal databases, and video files for signs that warn them of litigation risks.
An insurer today could also use this knowledge to come up with new products—a database of chemicals that might pose health dangers, for instance. This kind of commercialization of data is becoming more common. Often, the data chief is driving the train. Nearly two‑thirds of decision‑makers at organizations with CDOs say they are commercializing their data, compared with only a third without one, according to Forrester.
Although CDOs are still newcomers to the executive suite, their influence seems likely to grow along with the information they manage. Says Tony Cyriac, chief data and analytics officer at BMO Financial Group: “With data, you’re never done.”