Starting point
More and more people no longer start their research on a manufacturer’s website but in an AI assistant. They ask in full sentences - and so far get generic answers without current models, prices or availability. On the website itself, they have to click through filters and channels: configurator, new and used car listings, dealer search.
Solution
We built an app that makes the manufacturer’s product world available directly inside the AI assistant:
- Search in your own words instead of a filter obstacle course - matching models and variants in seconds.
- Models, variants and availability in one place: new cars, used cars and dealer stock instead of scattered channels.
- Straight to the next step: from the conversation into the configurator or to a nearby dealer.
- Real product data via connected interfaces - no answers from the model’s general knowledge.
What we focused on in the design
The interface lives inside the chat - it has to stay in the background and still guide clearly.
- Image before text. Every answer shows the vehicle large, with the starting price right below. People who scroll decide with their eyes first.
- Comparison in one view. Several model lines side by side, variants as rows with drivetrain and price. No switching between tabs, no mental notes.
- Every card ends with a next step. Configure, compare, find a dealer - clearly labelled, always in the same place.
- Make availability concrete. Vehicles with dealer and location, key data as short tags, plus a map of dealers who have the model in stock. The online question becomes a real appointment.
Result
- Less searching: from question to matching model in seconds.
- More relevance: models, variants and stock in one answer.
- A direct path from conversation to configuration and retail.
* Project: we owned the delivery · ● in operation: we run the solution today
