All transportation and logistics companies have data challenges, but not from the lack of it. The greater problem is what to do with all of it?
A speaker at the McLeod Software user conference in Dallas on Aug. 27, discussed artificial intelligence (AI) as the solution.
Tom Curee, vice president of strategic development at Kingsgate Logistics, shared how the company uses AI to stay on top of its ever-expanding data assets. Curee said the freight broker identified nine use cases for AI and has already developed and deployed it in six of these.
Specifically, he said the West Chester, Ohio-based freight broker uses machine learning in both commercial and internally developed applications to improve and automate tasks that normally would be done by humans.
Machine learning is a subset of AI and involves the use of advanced statistical models to identify complex patterns in data. Even more advanced is “deep learning,” he explained, which also is a subset of AI. Deep learning aggregates data from many different sources — called neural networks — and uses models that can learn on their own in an unsupervised fashion.
“I don’t see too many people doing crazy things with [deep learning],” he said, “but outside transportation there are some really interesting things happening.”
Machine learning caught Curee’s interest while he was the operations director at a 200-truck refrigerated fleet of owner-operators. His job was to fix its 133 percent annual driver turnover rate.
He brought that interest to Kingsgate Logistics and put it to work. Kingsgate arranges transportation for less-than-truckload and truckload shipments for shipper customers. Since April 2019, the company has expanded its technology team to 14 people, starting out with two business analysts and one network administrator. The company now has 90 employees total.
The reason for expanding the technology team is “because we see the need and commitment with what we have to do with regards to technology,” he said. “We are all feeling this pressure…to become more competitive.”
To give the audience an example of a technology for freight brokers that uses machine learning, Curee mentioned a new Top Match product from McLeod Software. Top Match is a module for McLeod’s PowerBroker transportation management system, he said, and uses a machine learning algorithm that predicts the optimal and best-chance carrier to offer a load to.
Data inputs used by the Top Match algorithm include carrier lane history, load board integration and margin data, he said. Curee said Kingsgate Logistics plans to use the product, but it already has been using machine learning in other areas.
Sales: “This is where we really wanted to do something different,” he said. The company uses an application that has a “best time to contact function” to identify who and when to contact. The recommendations come from using AI to analyze the details of customer activity for calls, email, website and social media use.
These insights feed into “prediction scores” used by Kingsgate to identify how likely a customer or prospect is to do business with the company, and what is most important to each client to prioritize sales calls.
For Curee, the “most exciting” application of AI is a platform it uses to automatically record and transcribe sales calls. Historically, a trainer would listen in on calls for 30 minutes to an hour at a time, for each salesperson. The platform is able to monitor all calls and identify every question, topic of discussion and other attributes to score calls and identify insights.
“We are starting to identify where are our best practices,” he said, and to model the practices of the company’s best salespeople. For example, the average call lasts 32 seconds, he said, but “we want customers to talk longer.”
Further insights reveal that its sales team spends about 40 percent of their time doing “small talk.” But small talk accounts for about two percent of calls for the most successful salespeople, who also spend 60 percent of their call time sharing market insights with customers, he said.
Marketing: The company uses a “SmartQ” feature of its social media platform from Zoho to publish posts at the optimal time for when each type of audience — customers, carriers, etc. — are the most active. “We create the content and AI does the work for determining when the content is going to hit,” he said.
Kingsgate also uses software to do “dynamic personalization” of its website. The software tracks the IPs of website visitors and through integration with the company’s CRM system knows what type of industry — flatbed, truckload, LTL, etc. — that visitors are part of.
The software knows if the visitor has come to the website looking for a flatbed broker, for example, and will tailor the website with content about its expertise in flatbed transportation. Likewise, the website will dynamically adapt to visitors from refrigerated and other industries.
IT support: Kingsgate has deployed an “automated reply assistant” to make its IT support team more efficient at responding to tickets. When a user needs IT support, they now interact with a pop-up messaging system that provides them with links to solve whatever problem the user is experiencing. The system continually learns as IT support monitors the automatic responses and grades them by clicking a “thumbs up or down” button, he said.
Kingsgate plans to use the same response system for customer service by feeding it data from the company’s PowerBroker database. The company could provide customers with immediate responses to load tracking questions and other inquiries.
Recruiting: Machine learning has made it possible for Kingsgate to more quickly find and hire the right candidates based on their job titles, skills, years of experience and other attributes that are the most important. The system continues to make recommendations as well.
“This is allowing us to refine our approach,” he said.
In these and other examples, Curee said the most important takeaway for AI is that it makes data actionable.
“[AI] is not always making decisions for you, but it is helping you be able to make the decisions you need to make,” he said.