Thanks to the so-called ‘Recommendation Engine’, users receive suggestions for individual vehicle options during the configuration process. The AI determines the suitable options from billions of possible combinations within a fraction of a second. Each recommendation is unique and tailored to the respective customer.
The ‘Recommendation Engine’ is based on several million data points from the Car Configurator. To implement it, more than 270 machine learning models were trained for specific markets and derivatives – so-called neural networks. These use algorithms to find specific patterns in the data, making it possible to predict options to provide the greatest possible benefit for each individual customer based on their own choices.
As a result, the recommendations are more than 90 percent accurate – meaning there is on average a 90 percent likelihood that a suggested feature will match a customer's interest. And because the ‘Recommendation Engine’ is self-optimising based on new data, this rate actually improves over time.
“By using artificial intelligence in the Porsche Car Configurator, we’re making the configuration process easier for our customers,” says Axel Berger, Project Manager in the Advanced Analytics and Smart Data department at Porsche.
“We’re using this technology to create a truly personalized online experience where we display relevant equipment options – no two users receive the same recommendations.”
Porsche has taken particular care to ensure data protection and anonymisation, so users’ privacy remains protected at all times.
Customers in six countries can currently take advantage of this offer: in Germany, China, USA, Great Britain, Switzerland and Taiwan; others will follow. They can access the virtual consultation via the ‘Recommendation Engine’ by clicking on a light bulb icon at the bottom of the page.
As part of its digitalisation strategy, Porsche is using AI in a way that is clearly geared towards customer benefit, pursuing the goal of inspiring customers by extending its unique product and brand experience to the online sphere.