Fast food advertising is so noisy and the US giants have the bucks to drown everyone else out. So whiteGREY had to use all its creative ingenuity (and technology) to give Australian chicken chain, Red Rooster, a BIG voice.
Actually, what it did was give Aussie tummies a big voice. whiteGREY and Red Rooster have launched crave_ware – an AI-powered technology solution that works out when your tummy is saying, “feed me” and orders Red Rooster for you.
The creative technology has been trained to distinguish tummy rumbles, aka borborygmi, from all those other bodily noises, such as burps and farts. And when it hears them, it places a customised meal order on behalf of the customer based on the intensity and duration of their tummy rumbles.
Chad Mackenize, chief creative officer, whiteGREY and Jono Casley, chief technology officer, whiteGREY took The Stable behind the scenes.
The Stable: Where on earth did the idea come from?
Chad Mackenzie: A fair question as it does feel like it could come out of the blue, but it’s an idea that comes straight from the brands purpose – to Satisfy Australia’s Chicken Cravings. Such a wonderfully honest purpose filled with spirit and intent. And one that wants ideas like this filled with fun and gravity. In essence, every brief we get from the Reds marketing team is asking for ideas of this shape, so to bring something to the table that made the team so excited to pursue was fantastic.
The Stable: Why a mad idea as a campaign?
Chad Mackenzie: I do love the term mad idea, but hopefully it’s also seen as an innovative idea, an idea that brings a genuine smile to your face, and an idea that will spark some great conversation. And conversation around innovation and fun is something that’s incredible valuable for Red Rooster. I guess the ability to shout louder than the big guys in the QSR sector isn’t an option for Reds, which is where creativity and audacity come into play. Crave_ware is an idea that brings these things to the table. It brings the brands body language into action and will make our fans lean in, and non-fans to sit up and smile. I mean, an idea that bypasses the traditional decision process and goes straight to the source of cravings is pretty cool, huh?
The Stable: How does the technology work?
Jono Casley: The underlying technology used for crave_ware heavily leveraged Google Tensorflow.js – an open source library for numerical computation and machine learning (AI). Crave is a pre-trained CNN (convolutional neural network) which can spot patterns in images for classification purposes. Ours has been exposed to hours of first party bodily audio data. We converted the audio data to an image-based data format known as a spectrogram, so the CNN is able to process the data. The model was then able to accurately classify user’s stomach rumbles and discriminate against other sounds. A user facing mobile app, running the trained model, receives and processes audio signals from a wireless listening device embedded in the user’s t-shirt. Once the correct conditions are met, a cloud-based automation tool processes the users order via the existing Red Rooster website.
How did you test it?
Jono Casley: Interestingly, training and testing are intrinsically linked when working with neural networks. We collated a large, anonymous, labelled dataset of bodily noises from our volunteers. Using supervised learning we exposed approximately 80% of the dataset for the network to train on, leaving the remaining 20% in reserve to test how accurately the network was able to classify the inputs and make any changes to the weight and biases between each layer in the network. This involves many training iterations/epochs to get our loss value as close to zero as possible, eventually we ended up with a fit model. On the more practical side, our volunteers wore the tech-enabled t-shirts in a variety of environments and performing various activities to ensure Crave was able to accurately identify rumbles and notify users when an order was ready to process by Red Rooster.
The Stable: Tell me about Isaac EatsALot and how you’ll spread crave_ware as far as you can
Chad Mackenzie: Red Rooster has a really vibrant, active social following on Instagram and TikTok. It’s where our community thrives and where we’ve seen and heard the passion and love for Red Rooster. Fuelling and engaging with this level of fandom is super important.
We engaged a range of influencers who will spread the word but kicked it off with an exclusive piece of content from the ever-enthusiastic, Issac_EatsALot. We asked Issac to join us on set and experience the technology firsthand – and of course satisfy his chicken cravings with some freshly fried chook from Red Rooster. He’s well known on the Australian food scene and has worked with Red Rooster in the past with existing and new menu items. He was ecstatic to get involved and spread the word of Red’s new wild tech.
We’ll continue to see a range of influencers engage with our content including blind reacts to our novel new AI tech.
Credits
Client: Red Rooster
Director of Marketing: Ashley Hughes
Marketing Manager: Jane Hillsdon
PR & Communications Manager: Stella Katsaros
Agency: whiteGREY
Chief Executive Officer: Lee Simpson
Chief Technology Officer: Jono Casley
Chief Creative Officer: Chad Mackenzie
Creative Director: TinTin (Ronojoy) Ghosh
Senior Creative: Kieran Adams
Senior Developer: Adam Shaw
Senior UI Designer: Jeremy Frouin
Managing Partner: Shane Holmes
Account Director: Liz Kerr
Strategy Director: Brooke Thompson
Strategist: Enid Sung
Head of Production: Colin Tuohy
Executive Producer, Director: Sean Ascroft
Production Assistant, Stylist & Art Department: Sarang McLachlan,
Post Production: whiteGREY
