Those who work with technology or marketing know that the artificial intelligence field is helping redefine many businesses by automating processes that previously demanded human interference. From fabricating products, through credit analysis, all the way to customer support, many things are changing with the evolution of the computer science field.
Robots have also been launched to the monitoring and social media analysis field. One of the most interesting applications nowadays is the automatic recognition of images in pictures and videos that are shared on platforms such as Instagram, Twitter, YouTube, and Facebook.
The automatic recognition of images came to solve an old problem: brands were invisible to monitoring in case they weren’t mentioned in the text. That means, if someone takes a picture with your product on Instagram bud doesn’t publish the hashtag #nameoftheproduct, that post is invisible to monitoring tools.
The automatic recognition of images opens up a series of possibilities for brands, which we call Listening 3.0. Know more about it:
Find out your share of image
Which brand is more related to Sustainability? Or to #junkfood? Or to #fitness? With the help of Buzzmonitor, we monitored the hashtags #snacks and #junkfood on Instagram to find out. These hashtags are used to indicate the consumption of sweet or salty snacks and treats, usually in moments when a routine is broken and the user treats himself with food. With the automatic image analysis we collected 50 thousand pictures associated to these hashtags and found out that 10,7% of the posts were about sodas, which is less than chocolates, which summed up to 16,2% of the detected brands. Within the sodas, Coca-Cola was the brand with the largest share of image, with 6,6% of the logos detected on the images. With the growth of Instagram, already being used by 45 million Brazilians, it is unforgiveable to ignore the share of image for interesting topics for your brand. The Brand Logo Detection technology allows you to analyze thousands of pictures faster than human analysts.
Identify spontaneous influencers
In 2016, the brazillian cachaça 51 made a placement on the Netflix show The Ranch. The cachaça appeared on the counter of Maggie’s Bar, a place where the characters of the show met. Using the new automatic image detection technology, it’s possible to look at thousands of videos and discover, in the sets of shows like The Ranch and on YouTube channels, products that are used for influencers that own those channels, even if the brand has never been mentioned on the text, but just like cachaça 51, are in the background. The influencer identification gets even more authenticity, since we discover people that share your brand’s products, they’re real consumers, but never mentioned it. That can help your company identify influencers more accurately, crossing the number of followers, bio description, and the use of your products through images.
Discover brands that are associated to your brand
When we looked at the pictures about #junkfood, we found out there were pictures where out Brand Logo Detection technology identified more than one brand in the same image. That’s interesting to find out consumption habits with combined products. Instead of having breakfast inside a house with a lot of people, as research institutes used to do in the past century for field research, you can simply monitor the hashtag #breakfast and find out which brands are being combined by the consumer. We discovered the combinations: Coca-cola and Heinz, and Nutella and Nido milk by looking at the #junkfood universe.
Interact with consumers of your brand (that didn’t say they were)
We can interact with people who haven’t mentioned the brand but are consumers of our products. That opens a new variety of possibilities for Social CRM. Combining that with Venues (geolocation monitoring on Instagram), we can offer a brewery the possibility to speak to a consumer who is in a bar in the neighborhood with a beer on the table.
Segment paid media
The sports brand Under Armour monitored runners that published pictures of their running shoes, as well as the competition’s running shoes. By doing that, the brand was able to segment paid media in a more accurate way.
And you, what do you intend to do with the automatic image monitoring?
Alessandro Barbosa Lima
CEO E.life Group