How Predictive Data Science is Driving Used Car Management Innovation | vAuto

“It’s like having the answers to the test before you take it.”

That’s how some dealers now regard the art and science of used car inventory management. The comment reflects how AI and predictive data science, when used to determine each used vehicle’s investment value potential, fundamentally changes how dealers acquire, appraise and price their used vehicles compared to the ways they’ve handled these mission critical decisions in the past.

“When you have a clear understanding of whether you’ve got an earner or a turner on Day 1, it changes how you manage each vehicle,” says Derek Hansen, vice president of operations for vAuto and Cox Automotive. “Furthermore, the understanding of each vehicle’s unique investment value helps shape strategic, ROI-focused objectives for all mission-critical aspects of used vehicle management.” 

This blog discusses how predictive data science is changing how dealers regard and treat individual used vehicles, and the solutions dealers use, to optimize used car profitability and sales at a time when the market is less forgiving and more profit-challenged than it has been in recent years. 

A Classic, Common “Test” for Every Used Vehicle

Historically, every dealer has faced the same test with every used vehicle: If I acquire it for $X today, how should I price to make the gross profit I expect to make when the vehicle sells?

Like individual students in a classroom, dealers answered the test question in their own way. Some price up from cost, marking up a vehicle high enough to leave room for discounting and gross. Some assess the retail market for the answer, picking a retail price point they believe offers the best opportunity to move the vehicle and deliver gross. Some use the retail market to price aggressively, choosing price points that position a vehicle to sell faster than the competition.

Despite the different methods dealers use as they take the test with each vehicle, two things

 are the same and true:

First, it takes time for dealers to know if they answered the test correctly. To be sure, you have some vehicles where a buyer is waiting for the used vehicle you just acquired. In most cases, however, it takes time for a vehicle at a certain price to get attention and interest from potential buyers that leads to an inquiry.  

Second, no dealer has ever aced the test with every vehicle. It’s a bold claim, but the fact that almost every dealer’s inventory includes a share of vehicles that should have retailed sooner suggests room for test-taking improvement. 

How vAuto Tapped Predictive Data Science to Helps Dealers

In the years prior to the 2020 pandemic, dealers were often selling record numbers of used vehicles and making less money. In some cases, used vehicle departments were losing money despite record-setting volumes. 

The trend of declining profitability led vAuto data scientists to attempt to answer a question: How can we help dealers know each used vehicle’s profit or ROI potential on Day 1, so they are better prepared to take the test with every used vehicle they appraise, acquire and price? 

The team developed advanced algorithms and applied them to a decade-worth of used vehicles and related transactions, and thousands of vehicles currently in dealers’ inventories. After much testing, the team was able to predict probability of sale of any vehicle at any price in a seven-day timeframe, with more than 90 percent certainty. 

With this understanding of pricing probability, the team began building predictive pricing models, which became the basis for a new inventory management system, ProfitTime GPS, that’s built to help dealers optimize the ROI for every vehicle using predictive data science to drive acquisition, appraising and pricing recommendations. 

The predictive data science behind ProfitTime GPS allows dealers to manage beyond what seems right in the moment—the equivalent of knowing the answers to a test before you take it. This unique capability provides dealers with one-of-a-kind foresight into the likely outcomes of every decision they make.

vAuto’s Investment-Minded Used Car Management Innovation

If you’ve got predictive data science-driven insights about every vehicle’s investment potential available on Day 1, you are now able to truly manage each used vehicle to achieve its optimal profitability or ROI outcome. This ability fundamentally changes how dealers regard used car inventory management: You effectively switch from managing cars or merchandise to managing your inventory like it’s an investment portfolio. 

At vAuto, we call this the Variable Management strategy. It encourages dealers to establish their used vehicle investment strategy—are you going for volume, gross or a blend of both?—and then calibrate ProfitTime GPS, and the predictive data science that drives it, to help execute the investment strategy as dealership teams acquire, appraise and price vehicles. 

How Predictive Data Science Enables ROI-minded Used Car Management
 
One of the key ways predictive data science helps dealers successfully take the used car test more effectively with each individual vehicle owes to them now knowing if a vehicle’s an earner, a turner or something in between on Day 1. 
At California-based Hello Auto Group, such investment value insights have helped GM Joe Seppa and his team adopt a used car management strategy through ProfitTime GPS that helps them manage each vehicle according to its risk and reward profile.
Seppa likes how ProfitTime GPS helps him bucket inventory by risk and opportunity, allowing used cars in each bucket to sell in distinct, investment-optimized timeframes to ensure each vehicle sells more efficiently and profitably in accord with its investment profile.
For example, Seppa notes, that “the cars you buy at auction are higher-risk cars. You need to buy them, but you need to manage them differently than any other inventory because it presents the highest risk.” 
Predictive data science insights with ProfitTime GPS also help Joshua Clinton, general sales manager at Cape Coral (FL) Chrysler, Dodge Jeep Ram, innovate his used car management strategy. For example, he now aims to align his used vehicle pricing to ProfitTime GPS’ data science-driven recommendations 90 percent of the time. 
“We’ve found that’s been a positive for our store,” Clinton says. “It helps us turn vehicles faster and make more money.”

Conclusion
Dealers who put a greater level of trust in AI and predictive data science are better able to navigate the current used vehicle market, where profitability continues to be challenged by limited used vehicle supply and volatility across distinct vehicle segments. Put another way, ProfitTime GPS and the predictive data science that enable the Variable Management strategy, can help you successfully take, and pass, the used vehicle test every time.