Predicting Dealer Buy-In Prices of Pre-Owned Cars

Selling a car can be a daunting experience. It takes time for the dealer to come up with a resale valuation, and there is always uncertainty over how accurate the figure is. The car dealers, on their part, need to provide a timely and accurate valuation taking into account the market conditions, and the condition, make and model of the car.

Quotz, a subsidiary of Singapore’s biggest online car auction platform sgCarMart, approached AI Singapore for help in addressing this business challenge. They were looking for an AI solution that could predict, more accurately, the dealer buy-in price for cars.

After consulting domain experts from Quotz, the AI Advisory team put some algorithms through several iterations before identifying the optimal algorithm and appropriate data engineering methods for the AI model.

Augmenting Existing Workflow
The AI model that was developed could ingest both numerical data (e.g. engine capacity) and non-numerical data (e.g. features of cars) and return predicted prices of the pre-owned cars. This enabled Quotz’s customer service officers to quickly generate more accurate initial valuations of used cars, thereby providing a better experience for their customers.

The solution has since been deployed and integrated into existing processes for Quotz’s customer services officers. “We are very encouraged by how AI can help with our workflow, complementing our existing services,” said Quotz’s general manager Glenn Ong.

With outreach to over 500 used car dealers and car owners all over Singapore, the incorporation of AI into Quotz’s business model has helped the company to enhance customer service and stay ahead in a highly-competitive industry.

For enquiries related to AI Advisory services, please email