xBerry Case studies DRNK App

DRNK App

augmented reality solution for the beverage industry

Watch the video

When are customers more eager to buy a product? When they realize that they receive more value from owning it than keeping their money. However, they often give up on a product when they encounter difficulties finding information about its benefits. How to engage customers in real-time and provide them with the information they may need while assuring a fun experience all at the same time?

Challenge

The beverage industry seems to be tricky to advertise. Due to the legal regulations in different countries, shop policies, and alcohol advertising restrictions on the Internet, our client needed to seek an innovative way to reach potential buyers of spirits. Based on a study of customer behavior, Omniaz came to us with an outstanding idea for a fresh way of approaching new customers, which at the same time would provide beverage retailers with a new communication channel.

We were up for using Augmented Reality in an extraordinary mobile application designed specifically for the beverage industry in the Asian market. The idea was based on the typical behavior of a client when making a purchase decision. It’s common that customers like to test a different type of wine or spirit, but they are not sure which one they might prefer.

The app would offer them fast access to information about wine, beer, and spirits they might want to purchase and present it all in a fun and interactive way by using augmented reality right before their very eyes via their mobile phones.

Goals

Together with Omniaz, we wanted to create DRNK APP - an AR mobile innovation for alcohol beverage advertising.

As the main application features, we needed to offer automatic detection of bottle labels, self-activated animation associated with a specific label, and display all the relevant information about the alcohol on the users’ screens.

We meticulously planned the fun factor for this solution to increase customer engagement and interest in a specific beverage and the entire spirit brand. This innovative approach helps customers uncover the value of the product in the decision-making process, and gives retailers a chance to distinguish their products from the competition.

Solution

We selected C ++ as the main language for the framework due to the requirement for fast data streaming (for speed of operation and effective data streaming).
We relied on OpenGL to create our own graphics engine rendering AR scenes running on devices with limited resources (to assure that our software works on various, even low-grade, phones).
We used Android Media SDK to interact with a graphics processing unit to achieve hardware-accelerated rendering.
We used TensorFlow for label recognition with deep learning.
We selected Vuforia for object tracking.

Results

The project had to hit close and non-movable deadlines, so it had to be built quickly but remain refactorable. We finalized the R&D part of the project in Oct 2019. After that, Omniaz tested the product over a 1.5-month-long campaign with a large retailer in Asia.

During this period, DRNK APP was tested in 24 stores for more than 500 labels per store. The application’s conversion rate hit 6%, and out of all 300 interested clients, 75% were new.