We’ve created the first algorithm that enables the analysis of gameplay on Twitch, automatically triggering the ad precisely when it detects that the character’s health points have dropped.

The core of the idea lays in the mechanism of data collection and analysis. It’s based on artificial intelligence which role was to teach computer vision application to recognize the energy level in the game. Then the machine learning model was used to trigger the ad at the right time. By creating a tool capable of recognizing in-game events on Twitch, we opened the doors to much more immersive communication for brands.

Danio – a nutritious cheese from Danone is an iconic brand in the dairy category in Poland. Its main brand asset is the Small Hunger – the oldest existing brand character on the market, whose role for several decades was to surprise and entertain the generations.

However, despite high awareness, the brand has been losing its relevance recently, especially for the younger generation who 1) doesn’t watch the ads 2) is very critical about marketing products 3) connects far better with influencers than brand heroes.

The task was simple – make the product and its brand character relevant for the youngsters.

To get their attention, we decided to implement the brand into one of their key passion points and consumption moments – gaming.

We decided to enter the gaming territory with a brand new tool that would allow us to distinguish ourselves from all the other advertisers on Twitch and connect with the main benefit of Danio cheese – the nutrition factor.

Because Danio cheese stands for regaining the energy, we decided to link the brand with the element that exists in every popular game – the energy bar. And own it.

None of the existing advertising formats allowed us to do so, therefore we created our own intelligent ad format, which thanks to AI, Computer Vision, and Machine Learning could monitor any game on Twitch and display the customized communication, whenever someone lacked the energy in the game.

The group we’re hoping to attract with our communication was extremely difficult – the youngsters basically think they’ve seen everything. We wanted to prove them wrong,

and as a medium to do that, we chose Twitch – the most popular streaming provider worldwide, with 73% of users under 35.

We didn’t want to base our strategy on specific partnerships or particular games. We planned to create a tool that would allow us to communicate in almost every game and totally own the player’s energy territory in the future.

Knowing that people who are watching the streams focus on the reactions of the streamers who are playing, we wanted to surprise these guys with highly relevant content, teasing them about the level of their energy exactly when it dropped.

The technology used in our tool is based on Artificial Intelligence, which role was to teach Computer Vision application to recognize the energy level in the game. Then the Machine Learning model was used to trigger Small Hunger for action when the energy ran out. Our tool was analyzing the game in real-time and was sending the request to trigger the ad to the media partner at the right moment.

The tool is capable of recognizing almost any event within the game and can become a new way for brands to customize their communication and deliver it at the right time.

It has been used for the first time in Fortnite during one week’s campaign that involved 128 hours of streaming.

Even the most popular streamers reacted to Small Hunger teasing them about their energy level.

Within one week of the campaign, we reached over 650 000 gamers.

We generated an Advertising Value Equivalent of 3x media spending.

We managed to stop the negative trend and gained 2,2x more positive mentions about the brand on Social Media.