Yuriy Demidko recently left DP Fox Ventures, a company that includes the Fox Motors dealership group in its portfolio, to work for what he calls a “tech forward, AI-related” company that is “really trying to make an impact on the auto industry as a whole.”
Fox Motors, based in Grand Rapids, Michigan, represents 61 brands in 50 locations, primarily automotive dealerships with a handful of powersports stores.
During multiple tenures at Fox, including his last as a senior vice president and chief information officer, Demidko, who also worked at Apple, was involved in many artificial intelligence initiatives.
He shared some of his favorite ways Fox Motors utilizes AI in a Zoom call with WardsAuto.
Creating a data lake and leveraging AI
The dealership group generates a massive amount of data. Under Demidko’s direction, Fox Motors over the past 15 years created a “data lake” integrating data from all the dealership group’s systems, including its dealer management system, its customer relations management software, its websites and “even the phone system,” he said.
The information was stored on a cloud-based platform which Fox Motors used for business intelligence insights. When AI emerged, Demidko layered an AI component over this business intelligence infrastructure to create what he called Fox Nexus AI.
Using Fox Nexus AI, Fox Motors employees can get information and answers to questions about any aspect of the group’s business.
More than that, however, the AI-enabled platform can generate reports including graphs and tables, incorporating the information and actionable recommendations.
For example, someone might ask Fox Nexus AI where expenditures have been trending for the past six years, identify where costs have risen “and then give me ways to attack the problem,” Demidko said.
Fox Nexus AI went into widespread use about eight months ago, and not just by the young, tech-savvy staff.
“It was adopted by people who have, you know, been thirty years in the industry,” Demidko said. Even those who previously shied away from technology “took it with both hands,” he said.
Getting special offers in sync
Auto manufacturers continually come out with special offers on specific models. Updating all of Fox Motors’ dealership websites was extremely time-consuming. Demidko created an AI tool he called Dealer Offers Management System, or DOMS, to update every dealership site with all new offers including images and compliance.
It’s easy to use. An employee fills out a form with the new manufacturer offers. DOMS generates all the disclaimers, makes sure all the proper fonts are applied, inserts a photo of the vehicle, then pushes it out on websites, social media, e-blasts and other marketing materials, Demidko said.
“This took something that was a 10- to 14-day bottleneck and turned it into seconds,” he said.
Analyzing commenter sentiment
Some groups pay consultants to analyze online comments. Demidko figured that was a waste of money.
“I was thinking about it, I’m like, ‘Well, what if we took all of the reviews our customers give us on the various Google, Facebook, all these platforms, put them in a database, and then point a large language model at it?’ So, I did that,” he said.
A large language model, or LLM, is a highly specialized subset of artificial intelligence.
Now, Fox Motors can generate a report with the top 10 topics about Fox Motors — negative or positive — people are saying about the group online. That creates a list of “low-hanging fruit that we can attack,” Demidko said.
One frequent complaint? Bumper-to-bumper warranties were a particular topic of dissatisfaction, he said. “People tend to get upset because they thought something was covered in bumper-to-bumper warranties and it was not,” Demidko said.
With that information in hand, Fox Motors finance managers told salespeople to devote extra time explaining what is covered in a bumper-to-bumper warranty.
Salespeople also let customers know that AI had identified a problem that owners of a particular vehicle model year consistently complained about, and which was not covered by the warranty.
The salesperson could then add, “Also, we want to make sure to let you know that, hey, for an extra $12 per month we have a service contract that will cover this XYZ problem,” Demidko said.
One the one hand, the salesperson is being transparent about the warranty coverage, he said. But they are also boosting the dealership’s bottom line “because you’re not quote-unquote selling, you’re providing a proactive fix to a problem that is likely to occur that other customers have brought up.”
AI-assisted leads, around the clock
In the retail auto industry in general, Demidko sees AI making the biggest impact for dealerships in handling top-of-funnel internet leads.
Fox Motors found that more than 38% of its customer engagement was happening after hours, representing lost business. “That’s a rather substantial metric,” Demidko said.
The group turned to software from Impel AI to handle those inquiries. Before implementing Impel, those same leads “would need to wait until business hours to receive a response,” he said.
That type of customer service AI tool is rapidly evolving, Demidko said. Rather than a uniform way of responding, some AI platforms now can handle leads coming in from multiple sources differently depending on how they are told to handle them, he said.
“So, that’s a step up from what it did, let’s say 12 to 16 months ago,” he said.
