SAN FRANCISCO – How big is big data? According to Andrew McAfee, it’s so big it’s about to outgrow the metric system.
McAfee, the principal research scientist at the Massachusetts Institute of Technology’s Centre for Digital
Business, told attendees at SAS Executive Forum conference that big data will soon outgrow the metric system. When analytics company Teradata was founded in 1979 it took its name from the Terabyte, then the largest amount of data you’d imagine working with. Soon we moved on to Petabytes, and now we’re in the Zetabye era, according to Cisco.
We’ve solve the issue, of course – McAfee jokes the leading contender for a new term is the Hellabyte – but the point is big data is, well, pretty big. And it’s getting bigger. Why does that matter though, and how can that help a business make better decisions?
As an example, McAfee points to the airline industry. It’s plagued with delays, but the airlines aren’t doing it on purpose. It’s about process. It’s important for an airline to know exactly when a plane will be wheels-down, so it can have the ground crew standing by. If they deploy ground the crew too soon, they’re sitting on the tarmac costing the airline money. If they deploy too late, your plane sits on the tarmac waiting for them to arrive. So they radio the pilot when he’s on approach to get an estimate – it’s often wrong, and the pilot is usually pretty busy flying the plane just then anyway.
“It’s a big data problem,” said McAfee, noting its one Passur Aerospace worked with an airline on a test case to try and solve. “You get the pilot out of the loop, and rely on historical data on weather, congestion patterns and airport information, with data from radar arrays. They did a much better job predicting arrival times and greatly reduced delays.”
If there’s going to be a challenge to wider big data adoption, McAfee said it’s not going to be technical, he’s confident vendors such as SAS will get the technology right. The challenge will be culture, and organizations adjusting to a new, data-driven decision making style.
“I’m not confident companies will effectively make the (internal) changes necessary to address the big
data challenge,” said McAfee.
The biggest enemy to this change, according to McAfee, is the HiPPO, or highest paid person’s opinion.
They’re the boss, they’ve probably gotten to that position by not taking too many chances and trusting their gut. They have to learn to go where the data wants to take them, even if their gut is saying otherwise.
“They need to switch to the geek style of decision making, but the HiPPOs won’t go quietly,” said McAfee. “They have a lot of hostility toward the geeks with their spreadsheets and analytics software.”
The data shows the geeks are more often right, and the HiPPOs are rarely right when they go against the data, but McAfee said he doesn’t want to make HiPPOs extinct. Rather, their role in the process needs to be modified.
“HiPPOs know what the challenges in an organization are. We need them to identify the challenges and then turn the geeks lose on them,” said McAfee. “The first role for an expert is to know what questions you should ask. You’ve got to know what you want the data to tell you. Then you need to make a decision based on that data.”