Mars ChocolateSnickers Hungerithm

The Internet gets a little angry when it’s hungry. So Snickers created the Hungerithm, a hunger-algorithm that monitors the online mood. The Hungerithm is like a stock exchange for Snickers chocolate bars. When anger goes up, Snickers prices goes down at every 7-Eleven stores. Prices are updated 144 times a day, and drop as low as 82% off the normal shelf price.

Clemenger BBDO Melbourne
2017, 2018
Vue.js, Webpack, Html5, Sass, ES6 JavaScript, Spritz.js, D3.js, GraphQL, Node.js, AWS services

The Hungerithm has been originally created for the Australian market, and as such, has been built to meet the local market needs. Following the original campaign success in 2016, Mars engaged with Clemenger BBDO for a global rollout to more than 50 millions customers across US.

In that context, I’ve been in charge of the technical approach for the rebuild, while leading a small team of developers, to deliver on a highly scalable and highly available solution for both the web platform and the mood analysis algorithm.

Winner of 6 Gold Cannes Lions, plus awarded on Webby Awards.

Technical breakdown

Built upon Google Deep Learning models for Natural Language, the Hungerithm determine sentiment by scanning over 14,000 social posts a day. It even understand slang, sarcasm, emoji’s and differences in context.

To redeem, users simply visit the Hungerithm mobile site and click the “Get A Snickers” button to receive an instant 7-Eleven coupon right on their phone.

“The Goo” is the name of the 3 animated chocolate waves built to bring some life and movement to the Hungerithm. It synchronise in real-time according the mood of the internet.

For content moderation purpose, we’ve also built a custom administration interface that controls and update the Front-end site in real-time.

The infrastructure fully rely on AWS services. An example of it is the data API, built with Node.js and GraphQL inside a Lambda function.