23 Apr
2019

The Audi e-tron Machine Learning Showcase

Although rapidly advancing, modern retail often fails to meet modern expectations. This is particularly true in the automotive industry, where 99.6% of customers say that the current car-buying process is not ideal (Autotrader). At the heart of much of this frustration is the lack of personalisation (Segment) and consistency (SalesForce), which creates a disconnect between a customer's online and in-store brand interactions.

The Audi e-tron machine learning showcase, developed for CES 2019, addresses these issues. Building on ZeroLight's earlier work with the ICC and connected customer journey demonstrations, it integrates all stages of the car-buying process to deliver a unique and consistent experience for each customer. The showcase enhances both the home and dealership stages with high-quality digital content and intelligent recommendations. This is done using anonymised customer behaviour data, thus creating a more consistent and personal experience.



"With this showcase, we have demonstrated how retail experiences can be personalised to every individual customer, at scale. With each stage of the purchase journey linked together, both home and in-store interactions can now be shaped around customers' needs, driving value through relevancy throughout."

François de Bodinat, CMO, ZeroLight

A hybrid of advanced technologies

This showcase uses three key aspects of ZeroLight's platform to create a retail experience beyond anything currently available.

1. Consistent visualisation solutions

Linking every stage of the customer's journey using the same 3D master digital twin asset across each output, ZeroLight's omnichannel platform delivers a seamless end-to-end purchase experience.

2. Real-time behaviour analysis and next-best action suggestion

ZeroLight's platform tracks and analyses anonymous real-time user data from every interaction. Machine learning algorithms then use these data to suggest the most relevant next action to each customer and the next-best offer for dealers.

3. Automated content creation

Images, videos, and interactive microsites are automatically generated using ZeroLight's cloud-based content creation platform, SpotLight™. These assets can be integrated into any advertising, CRM, sales, or communications campaigns and customer give-aways.

Maximising engagement at the home and the dealership

Using the key technologies outlined above, the e-tron Machine Learning Showcase encourages customer engagement both at home and in the dealership through a range of innovative features.

At home

A media-rich online advert that is personalised to suit any user persona (which can be built from a dedicated campaign or from CRM information) is automatically generated and displayed to the customer. This advert takes them to the e-tron configuration page where they can use real-time 3D visualisation to customise and explore a digital twin of their car. Throughout this process, bespoke machine learning algorithms analyse the customer's every interaction with the configurator to suggest individually relevant next-best actions that are designed to encourage engagement and conversion.

Once they have picked their favourite options, customers can see how unique their final car configuration is by comparing it with different customers' choices using aggregated data collected across other sessions. They can then enjoy their creation in a range of automatically generated beauty shots and videos before using a custom AR app to see how it looks in their real-world surroundings.


In the dealership

After completing the home stage, the customer loads their pre-saved configuration at the dealership where they can use a 4K video wall or virtual reality to review their choices with a dealer. Meanwhile, the dealer can use the tracked behavioural data to map the customer's online journey and gain a more accurate understanding of their needs. With this information, the dealer can suggest individually relevant upgrades and options to their customer based on next-best offer and recommendations tools driven by machine learning.



Once the configuration is finalised, the dealer can generate a unique microsite based on their customer's car. Created using SpotLight's automated content production capabilities, the microsite features images, videos, and interactive elements that showcase the car in a range of environments, thus replacing generic and static traditional brochures. The customer can share this content with friends online and download the personalised images and videos, encouraging distribution and engagement on social media. In the meantime, the dealer will be able to understand and analyse how their customers interact with their new interactive microsite.

ClientAudi
ProjectMachine Learning, Retail