Skip to main content

If You’re Drowning in Data, You’re Doing it Wrong

Loyalty programs. Store credit cards. Web searches. Buying histories. The apparel industry has woken up to the need to amass as much information about consumers as possible in order to design the ideal product, target the right shoppers and build appealing promotions. But none of these things will prove fruitful if we don’t know what to do with all of the facts and figures.

The saying “data is the new oil” is meant to convey the value of data. But the fact is, data on its own is worthless. And in that way, it is, in fact, just like oil.

Crude oil by itself doesn’t power your car. It first has to be refined into something useful. The same is true of data.

Raw numbers alone aren’t useful—especially in the quantities that are available today. Keep in mind, the average internet user generates 1.5 gigabytes of traffic per day. And soon self driving cars will create 4 terabytes in the same time period.

Data is everywhere but you must refine and interpret it. This was the message at the Intel SHIFT 2017 conference Tuesday. From Intel executives to their partners in fields as disparate as entertainment and health care, each underscored the need for companies to understand their needs and work toward a specific goal.

Be outcome oriented

Lisa Davis, vice president and general manager for Intel’s IT Transformation, Enterprise and Government Data Center group, said analytics and AI can help boost speed and agility, but companies must keep three things in mind.

First: “Business outcomes must drive your data strategy.”

Put another way by Davis’ colleague, Bob Rogers, chief data scientist for IT Transformation, Intel’s Data Center Group, said, “Find ways to tie your analytics to value or at least have clear metrics so people can understand that your metrics have moved the ball forward.”

Related Stories

While you may live by your time and action calendars, no one will die if your next delivery is a day late. At hospitals like Montefiore Medical Center, however, patients are battling life or death and some deteriorate quickly for no obvious reason. Michelle Ng Gong, MD, MS, associate director for Academic Affairs, and her team use machine learning and predictive analytics to aid physicians and provide better care.

“What we needed was something that would develop a 24 hour, 7 days a week surveillance that’s never tired or sleep deprived and always consistent,” she said, explaining how the technology identifies patterns in patients that are then used to alert doctors to potential problems.

Ultimately, the data is valuable because it serves a specific purpose, Gong said. “It’s only useful if I implement it in the right way to be able to help with the workflow of the clinician to get the job done. Do we have a common goal? Do we have a common purpose? And do we have a way of using that in such a way to help with this common goal and purpose?”

Remove the silos

Davis’ second key takeaway was analytics must be “democratized,” meaning it needs to be easy to understand and accessible to more people throughout an organization. “The traditional IT approach to data has been that of gatekeeper—you need to come through IT to access the data. But that’s not fast enough. We slow the business down. A lot of that was because the tools weren’t intuitive. That’s no longer the case,” she said.

Basically, your organization can’t continue to work in silos, where IT handles data and everyone else is off working independently. Davis said the data must be defined the same so if you’re going to look at information about your in-store shoppers and your online customers, for instance, there’s a common language that allows for an apples to apples comparison. This requires coordination across the whole organization. “Implementing system changes is hard,” she said. “In fact we say, it’s not the technology that’s hard. It’s the change management that makes it hard.”

Getting the information into the hands of those who can react to it is a big part of what Sean Petterson focuses on.

The founder and CEO of Strong Arm Technologies is out to assess and limit workers’ injury risk in industrial environments like warehouses. By using sensors on the body, his team is able to capture data at a rate of 13 times a second while what he calls “industrial athletes” are completing tasks like hauling boxes and picking and packing. “The amount of impact that has on your body is absolutely brutal. For context, they’re spending $250 billion a year on compensation claims just in the United States,” he said.

“We’re backed by science and generating all sorts of amazing algorithms but it’s kind of useless if it’s not simple,” Petterson noted. “We provide [the workers] a fantasy football-style draft dashboard on their performance. Then [it’s onto] the facility managers who make genuine insights into what they can do to make their jobs better.”

Look to the future

Finally, Davis said forget slapping Band-Aids on your current system. You must be forward thinking to keep pace with innovation. “You can’t win in this environment with legacy infrastructure,” she said. “We have to optimize the infrastructure to leverage the power of the data. It’s a holistic transformation.”

No matter what your business goals are—sell more, increase speed to market, understand customer behavior—data can likely help you achieve it. But that’s only if you can focus on your goals and apply the information accordingly. It’s a big task, Rogers said, but it’s possible you won’t have to start from scratch.

“The key is to learn from the best—not only in your own industry,” he said. “Someone has solved a problem that maps to a problem you have.”