Finding ROI in AI

September 18, 2023
In the News

As originally featured in Automotive News (September 18, 2023).

Author: Kelly Taylor

Dealers can effortlessly turn data into decisions and complete time-consuming tasks in seconds

ARTIFICIAL INTELLIGENCE is playing an increasing role in Canadian automotive retailing, helping sales staff create compelling advertisements, automating choices on which cars to advertise and expediting extended-service-contract claims, among other tasks.

Nikhil Chugh, country manager for advertising-technology provider Cartelux Canada, said that while the first use might have involved a platform such as ChatGPT helping write vehicle listings, AI’s influence continues to grow.

AI is helping brands sell cars more quickly after entering inventory, Chugh said, citing statistics that show it can take 15 days or more to sell a vehicle. It takes that long for dealers and automakers to work with their media agencies on the platforms to use and the audiences to market to, he said.

“With the platform we run, all of this is done in 60 seconds,” he said. Cartelux uses 300 million data points to analyze such decisions and provide solutions, Chugh said. This is aggregate data — such as the number of searches for eight-seat vehicles with all-wheel drive, or two-seat sports cars — mined through online analytics software. Crunching that data helps the Cartelux platform advise customers which vehicles are in demand and by which demographic groups.

Brent Sayles, president of Applewood Auto Group, is an early adopter of AI. He uses it in a variety of forms to write vehicle listings, to help sales staff who struggle with grammar respond to customer texts and emails, craft operating procedures and target sales promotions to the right customer at the right time.

Sayles, whose British Columbia-group includes 11 dealerships from Port Hardy at the north end of Vancouver Island, to Surrey, just east of Vancouver, said he has subscribed to 30 AI applications. He is more interested in building his own in-house AI solutions, often collaborating with vendors rather than buying off-the-shelf applications.

Examples include using Jasper AI to create 35 job descriptions in one day, a task that can often take hours per job description.

“I don’t think I could do 35 job descriptions in a year,” he said.

Sayles encountered a humorous example in an email from his accounting department, written in a tone that got his hackles up.

“They didn’t mean anything by it, but I thought, maybe we can use AI to create emails that don’t make you want to walk over there and pick a fight with someone,” he said.

Sayles’ next project is developing a system that uses the various points of data that his group already has to target customers with timely promotions, such as when service appointments suggest that it’s time for new tires.

“We could send them a note that says they might want to consider replacement — and, oh, by the way, here’s an offer for buy three, get the fourth free.”


For, which serves dealerships of all sizes, the just-launched Upgrade Optimizer AI performs a similar function automatically, said Ian MacDonald, the company’s chief marketing officer.

Autotrader publishes online listings of new and used vehicles and for added fees can boost certain vehicles to increase their visibility on the site. MacDonald said Upgrade Optimizer will, given a dealer’s budget, boost certain vehicles depending on a large number of data points, such as search popularity, as well as when a rise occurs in the number of searches — based on features such as seat counts, all-wheel drive or engine style — that match a vehicle in inventory.

Freeing up that time allows Autotrader and the dealer to concentrate on what matters most to the retailer: How much did it cost, and how much did it help?

“That’s part of the beauty of it,” McDonald said, “in that we can begin to have conversations … that’s more relatable to an advertiser — in this case, a dealer — who can say, ‘This is how much I want to invest, and this is the kind of ROI [return on investment] I’m looking for.”

LGM Financial, which provides extended-service contracts either directly or behind the scenes at dealerships, recently began using AI to expedite claims, with many approved in as little as five seconds.

Automating simpler claims frees staff to adjudicate more complex ones, said Scott Rutherford, LGM’s executive vice-president of technology.

At the retail level, Rutherford said, LGM also provides a “recommendation engine” designed to simplify the purchase of extended-service contracts and ancillary products. LGM provides service contracts sold through dealerships as if they were sold by the brand, he said. Such contracts have nine categories, but with thousands of options.

By asking a series of questions, Rutherford said, the recommendation engine helps narrow choices for customers. — ANC