New XPO Connect tool another step in substituting machine learning for human intervention

In a tight market for carrier capacity, keeping carriers engaged is critical. In rolling out an extension to its XPO Connect digital freight marketplace, and the XPO mobile app that drives it, XPO Logistics (NYSE: XPO) is seeking to be able to give a carrier more than one bite at the freight apple in case its initial bid for a load misses the mark.

According to XPO, the program works like this: XPO and its customer post a load at a certain rate. A carrier either agrees to haul at that rate or counters with a bid that is, say, $300 higher. XPO’s technology, using machine learning and other analytic tools, analyzes all the variables and reverts to the carrier with what it believes to be a fair offer, which could very well be its original offer. The carrier either walks away and XPO looks elsewhere for capacity, or, based on how its fleet (or in the case of an owner-operator, its truck) would be positioned upon completion of the head haul, agrees to accept the XPO counter. If the latter occurs, the load becomes immediately visible to the fleet or the driver for haulage.

Under the existing technology, the counter offer would have been made by a human being. Now, it’s the machine learning tools that will make the counter.

There is a larger purpose to be served beyond pulling down a more favorable rate for the shipper, according to Mario Harik, XPO’s chief information officer. First off, the process is digitally executed, meaning no human hands—or thought—enters the picture. Second, and maybe most important, it keeps the carrier in XPOs system, meaning the carrier stays engaged, Harik said in an interview.

Rather than forfeiting the load outright, the carrier or driver can assess potential backhaul opportunities (XPOs system can provide those or the driver can input its preferences when it signs up for the service) and discover that it makes sense to accept the revised bid if a return load is nearby. As Harik sees it, the software opens up more capacity for its customers, while giving the carrier a second chance to take a headhaul load—where most of the ones is made—

if it is rational to do so. Even in tight markets, carriers and drivers have backhaul needs, Harik reasons.

In tight or slack markets, a carrier’s top priority is asset utilization. That is what the software is designed to achieve, according to Harik. “We want to make sure our carriers benefit from continuous moves to maximize revenue,” he said.