Big change in how WeBuyCars buys and sells cars in South Africa
WeBuyCars, one of South Africa’s largest vehicle trading platforms, is making a significant change in how it buys and sells vehicles by placing artificial intelligence (AI) at the centre of its operations.
According to chief digital officer Wynand Beukes, the company is currently running two major AI systems internally. “We’ve got two big AI drivers in WeBuyCars, called Blue and Orange,” he said
“Orange is the customer-facing large language model (LLM), and that’s on our website. It also includes AI functionality used by our internal teams.”
“Blue is a collection of machine learning models that hold our pricing and vehicle information, based on historical data, purchase and sale history, and market trends.”
These AI models are updated weekly and used to guide decision-making, particularly pricing, which is essential in the vehicle trade business.
Since late 2022, AI has become more mainstream, mainly due to the introduction of tools like OpenAI’s ChatGPT.
WeBuyCars said the technology is already being put to use. “Blue has bought just over 2,800 cars autonomously, without any human pricing involved. We’re scaling that up as we go,” Beukes said.
These systems are designed to handle repetitive and high-volume tasks efficiently. Beukes explained that agentic AI is able to break a job down into smaller steps, plan how to do it, and work through those steps without human intervention.
The shift to AI is the latest development in WeBuyCars’ digital overhaul. The company began over two decades ago as a small operation run by brothers Dirk and Faan van der Walt.
In the early days, most processes were manual. When Beukes joined the business 17 years in, even core functions such as inventory, sales, and finances were still being run through basic spreadsheets.
“Everything was managed on a Google Sheet. And they sold over 2,000 vehicles this way a month. But it had reached breaking point. We couldn’t just throw more people at the problem. We knew we had to digitalise,” said Beukes.
The next step
He noted that at the time, the company faced a choice: adopt a standard enterprise resource planning (ERP) system or build its own.
“2,000 cars is a lot of stock to manage, and it’s difficult to do that on a Google Sheet. We decided to build our own software from the ground up.”
That decision laid the groundwork for the system that runs the business today. It manages all aspects of the company’s operations, from sales and pricing to customer leads and marketing.
WeBuyCars now operates 17 large vehicle showrooms, known as “vehicle supermarkets”, and nearly 100 smaller “buying pods” across the country.
Because it controls its own software and data, the company is able to test new ideas and features quickly.
The company’s marketing division also plays a large role in feeding the AI systems. Beukes said the business generates many of its sales and purchase leads through its website and social media platforms.
AI helps process those leads, allowing human staff to focus on the more unusual or complex cases.
“If we can automate a certain percentage of our lead or buy-lead management, we can handle more leads with the same number of people,” explained Beukes. “Then we let the humans focus on the edge cases.”
Not all vehicles are suitable for automated pricing, however.
“There’s no reason a human has to price a Polo Vivo, because the volumes are so high. But in fringe cases, like a 1974 Mercedes-Benz, that’s difficult to price. It needs human understanding.”
Beukes noted that the company isn’t aiming for full automation, but rather for a system where technology handles the bulk of routine tasks and human expertise is reserved for where it’s most needed.
He said AI is simply the next step in making the business more efficient, scalable, and responsive to customers.
Loading ...