AT&T Executive Delivers Big Insights on Big Data at Workshop

Connecting power-of-10 terms such as petabyte, exabyte and zettabyte to the everyday realities of running his business, John Donovan, senior executive vice president of technology and operations for AT&T, gave the industry keynote address at the 25th annual Workshop on Information Technology and Systems (WITS) hosted by the Naveen Jindal School of Management.

In introducing Donovan, Dr. Varghese Jacob, vice dean of the Jindal School and co-chair of the WITS gathering, acknowledged AT&T as one of the school’s “strong partners over the years.”

The audience of information technology professors and researchers listened to Donovan describe the impact big data will have in an increasingly connected world and AT&T’s big data initiative.

“For us, at our scale at AT&T, the use of data is a game changer. It’s nothing short of the biggest thing that we’re doing because it primarily transforms how to make decisions and when we make decisions, how things get done,” Donovan said.

Massive scale was a recurring theme as Donovan talked about the digital landscape and worked through some AT&T specifics in his keynote, “Big Implications of Big Data.”

As part of a data-intensive industry, AT&T’s fundamental business is delivering data on behalf of customers, Donovan said. He said that AT&T’s network carries more than 100 petabytes (1 million gigabytes) of data every day, with more than an exabyte (1 billion gigabytes) in storage. He also said that by the end of 2016, industrywide data delivery is expected to exceed a zettabyte (1 trillion gigabytes).

“If you think about that as a business opportunity, you have a massively scaled network that has opportunities in everything from how you handle your customers to how you’re repairing and managing the network,” Donovan said.

For most companies, he said, a problem is that data is siloed, not readily available to use, and it is often viewed as business intelligence that only brilliant data scientists can use to come up with blazing insights. Donovan said that AT&T thinks differently about big data. The Dallas-based company standardized processes to make analyzing complex data easier for employees, not just data scientists.

To arrive at those processes, the company had to enhance the usability and capability of data by inventing and building what didn’t exist — “a unicorn” in Donovan’s words — of function platforms to extract, abstract, synthesize, embed and do even more with data to solve complex technical problems.

He said AT&T spent more than $500 million in the last three years to do this. But it also crossed more than $500 million in cumulative benefits last year.

“So for us the problem of getting started was the problem,” Donovan said. “It turns out that this is not only good science, it’s really good business as well.”

He cited how AT&T saved money by analyzing telematics data from its 100,000-vehicle fleet. By tracking vehicles’ mileage and use, AT&T found it could reliably predict when alternators and batteries would fail. Running some gaming tests, data scientists discovered the company could reduce total costs by changing alternators and batteries before they failed.

“The impact on the business is a small ripple from the standpoint of what you do every day,” Donovan said. “But you have profound benefits, approximately $1.5 million in this case.”

Donovan said AT&T steadily tries to enhance the business intelligence field internally and continue advanced work to move from descriptive to prescriptive analytics and beyond.

“If you have information, you can create insight that allows hindsight — that has value because it allows change, and you can produce a different outcome the next time,” he said. “If I can explain what happened, maybe I can prevent what’s going to happen. That’s the next step of the journey and so on until we get to ultimate optimization — where events don’t occur because we proactively managed them with prevention. That’s really what the network is all about.”