by Clint Boulton

Dealer Tire gains traction with data science

Apr 08, 2019
AnalyticsData ScienceIT Leadership

Dealer Tire is predicting when tires and other automotive parts will need replacing, an approach it says will help dealership and manufacturer partners generate more revenue.

blazing speed leaving a glowing trail along a road
Credit: Thinkstock

In 2016, Dealer Tire executives posed a salient question to the 100-year-old company’s fledgling data science team: Can we predict when each consumer will need tires? The answer — critical for a company that must get to customers before they buy tires somewhere else — was a resounding yes.

“It was the first ‘a-ha’ moment,” Chris Schron, Dealer Tire’s director of data science, tells “We realized if we could build some predictive models to answer that question, it could really add value.”

That has paved the way for Dealer Tire, a distributor of tires and other automotive parts, to leverage analytics to create new data products and consulting services it could sell to its dealerships and auto-manufacturer partners. The value-added distributor is betting that big data will yield big money.

Strategies that leverage troves of transactional and other data to boost revenues abound. To fulfill these big data goals, companies are investing in analytics projects, whose core tools include software that can clean up, organize and model data for cultivating business insights. Worldwide revenues for big data and business analytics software will reach $260 billion in 2022 with a compound annual growth rate of 12 percent from 2017 to 2022, according to market research from IDC.

From descriptive to predictive models

To appreciate Dealer Tire’s big data aspirations, it helps to consider its past. In 2015, Dealer Tire brought in a third-party firm to gauge its use of data analytics; it didn’t like the results. Years spent collecting historical data about tire treads and other automotive parts meant the company was long on the so-called descriptive analytics that paints a picture of what has already happened.

But it was woefully short on analytics for anticipating customer trends and behaviors, says Schron, who the company hired in 2015 to address these gaps. Dealer Tire shifted its focus to predictive analytics, to help forecast what is likely to happen, and prescriptive analytics, to help determine what should be done.

Schron assembled a small team of junior and senior data scientists to build an analytics platform that could help analyze, among other concerns, when customers would need new tires — a seemingly basic need that can have bigger financial ramifications, as people who buy tires at a dealership are 2.7 times more likely to return to the dealership for service and are 1.3 times more likely to purchase a new vehicle from the dealership, according to Dealer Tire.

Creating a “smart tire,” then, can increase business for tire makers, dealerships and Dealer Tire itself. “Our whole business model is predicated on working with our partners to identify new opportunities that will increase sales and customer satisfaction,” Schron says. “If we could know these things in advance, we could capitalize and benefit.”

Traditionally, companies relied on “tire timers,” essentially estimating tire wear based on universal average miles. For example, the average tire wears out at 33,000 miles, with drivers logging around 10,000 miles per year. Yet all tire treads are not created equal and every driver does not average the same amount. Viewing all drivers as the “average” driver typically results in some consumers being reached by Dealer Tire and its dealership and manufacturer partners too early or too late.

Schron’s team merged data from car manufacturers and dealerships to model individualized tire wear to identify the prime time to drive each consumer to their dealership. The resulting Tire Trigger application features models that deliver individualized predictions. For example, one model predicts the mileage at which a customer will need replacement tires, while another model predicts the number of miles per day that customer will be driving.

These models include sub-models that ensure the most accurate predictions to accurately predict tread depth of each customer’s car. Dealer Tire aggregated the models to get a replacement date, providing manufacturers with a list of every individual that has likely hit a specific tire tread depth, based on their vehicle and driving patterns. The dealership can use that information to determine when to notify their customer that it’s time for a new tire.

Tire Trigger has regularly delivered a 16 percent increase in raw conversion rates over a control group (a portion of customers who did not receive any marketing suggesting that they buy new tires), which corresponds to a 21 percent increase in revenue. Dealer Tire expects to scale the app to more dealerships and manufacturers in 2019.

A data science startup within a tire distributor

Dealer Tire’s data science work is resonating in the industry, with manufacturers turning to the company for help with anomaly detection and other use cases, Schron says. The work has also emboldened his data science team. Dealer Tire is currently putting Service Advisor Coach, a performance tool that provides staff with training, education and other tips on how to better sell car parts, through its paces.

Dealer Tire is also developing predictive models to help anticipate demand for different parts and services, as well as customer churn. Ideally, if a dealership can touch base with a customer at the right moment, they can head off a defection, Schron says. It is also selling, via consulting engagements, its predictive and prescriptive modeling expertise to help car manufacturers bolster their sales.

That the data science team is building analytics products to license to partners means it is operating like a “startup within Dealer Tire,” Schron says.

Dealer Tire combines raw data from multiple data sources and munges and transforms it for modeling with a software platform from Domino Data Lab. The data science team also uses Domino to share, manage and deploy code. Domino’s platform is user friendly enough for Schron’s junior data engineers that it has enabled the company to rapidly scale its analytics efforts.

It’s a far cry from the data science team’s previous approach of “duct-taping” analytics projects by pulling down SQL queries from Oracle databases and running them against the company’s data lake, which impeded data modeling, Schron says.

Data suggests Dealer Tire is on to something. By 2020, analytics that can help companies predict outcomes and prescribe courses of action will attract 40 percent of enterprises’ new investment in business intelligence and analytics software, according to Gartner research.

And Schron says the new approaches have helped Dealer Tire evolve into a more data-driven culture. “We look to continue that and embed that thinking across the organization,” Schron says.