Maersk spends $1 billion per year repositioning empty containers. Local managers tend to order too many containers to make sure they have enough on hand.

These are two of the conclusions drawn by a new group within Maersk devoted to data analytics.

It’s a problem no doubt faced by all large carriers, but even more so by the largest container carrier, Maersk, a company that carries 15 percent of the world’s GDP.

The shipping industry has historically depended on global economic growth to grow demand for its services. With the world economy slowing and global trade in the doldrums, carriers like Maersk can’t expect the world economy to drive the demand that will grow its business.

The container shipping industry currently suffers from a bad case of overcapacity. “The problem with overcapacity,” explained Jan Voetmann, who heads Maersk’s data analytics group, “is that it drives prices down. That leads carriers to try to become more efficient by doing things like deploying larger vessels. That creates even more overcapacity.”

The effects of Maersk’s bottom line are telling. The Copenhagen-based carrier’s parent company, A.P. Moller-Maersk reported losses of $107 million for the first half of this year.

On top of all that, companies like Amazon are entering the logistics space. “We see Amazon as an example of our new competition,” said Voetmann, who spoke at the Teradata Partners conference in Atlanta on Monday.

Maersk has decided to tackle the problems of efficiency and growth with by applying insights derived from data analytics to its business problems. Maersk’s pre-existing fact-driven culture, according to Voetmann, predisposes the organization to that kind of approach.

The new effort started in the finance department where it had the most willing sponsor. It has since expanded to the commercial and operations areas where the company can identify opportunities and act on them. “That’s old hat in the B2C world, but it’s something new for B2B,” said Voetmann.

Much of these new efforts revolve around studying the movement of containers. Studies conducted thus far found an excess of empty container movements. Because global trade is asymmetrical some movements of empties is necessary. But the analytics team tracked instances in which containers were moved five times without carrying a shipment, sometimes ending up where they started.

One container was ferried back and forth empty across the Pacific between Asia and the United States 20 times. When queried about this, a Maersk port manager in Asia acknowledged the situation and said he kept on sending the container back because he didn’t like the looks of it.

Another data project is looking at container ordering and the rightsizing of those orders. Local managers tend to order too many containers to make sure they have enough on hand, especially if they were burned in the past by getting caught short. The goal is to use data analytics and forecasting to position the correct number of containers when and where they are needed.

This new phase of Maersk’s analytics efforts are relatively new and Voetmann was unwilling to report on any specific results. He did say that his group intends to ramp up to the point where they’ll have something new to roll out every month.