Coca-Cola Targeting Ads According To Your Instagram Posts

What Happened
In an innovative twist of using social listening for targeting, Coca-Cola’s new ad campaign is targeting people the photos they shared on Facebook and Instagram, instead of the text they posted. The idea is to gauge user interests based on the images they share on social media and serve up relevant messages across other digital channels. For example, if you posted Instagram photo of the picnic you had over the weekend, and there happen to be a jug of iced tea somewhere in the mix, Gold Peak, a Coca-Cola-owned ice tea beverage brand, could target you with ads while you read an article online or checked the weather on a weather app.

What Brands Need To Do
We have entered an image-first era on social media where many people have chosen to prioritize sharing photos and videos over texts. After all, a picture is worth a thousand words. And thanks to the recent advances in image recognition and machine learning technologies, this trend provides brand marketers with a new opportunity to discern the interests of their audience and get a better understanding of contexts they are in. More brands should start expanding their method of customer data acquisition to incorporate image-based data collection and leveraging it to optimize the relevance of their ads.  

For more information on how brands may tap into the transformative power that machine learning will bring to marketing, please check out the Augmented Intelligence section of our Outlook 2017.


Source: Digiday

L’Oreal Targets Makeup Based On Your Outfit

We’ve seen product recommendations highly targeted online, but not so offline. Glamour is trying to bring product recommendations OOH, piloting kiosks in subways which recommend products after detecting the colors of your outfit. While the targeting parameters are novel, it remains to be seen if shoppers open to purchases in these pop-up style environments.

Apple Patents New Image Recognition Device Lock

Apple has filed a new patent application to cover a proposed method of unlocking digital devices using image identification.  The process would work by displaying an image or images from a user’s library and using either text or voice input, having the user identify the subject of the image exactly based on previous input.  This could be adapted into a highly secure method of device locking with the use of multiple images and highly specific identifier keywords, and could be a draw for security conscious users.