US company Ditto Labs has created software that identifies brand logos on images posted on social media sites such as Twitter and Instagram.
Using social media to decipher what shoppers think of a brand is difficult – there’s a huge amount of data, and it’s not exactly structured.
Ditto Labs has come up with a way of using photos posted on different networks to aid brands’ understanding of how they are regarded.
It has created image-recognition software that identifies brand logos on pictures, allowing businesses to be alerted to pictures that feature their brands that are posted even if they don’t mention them by name.
Any pictures uploaded onto social media websites such as Twitter, Instagram, Tumblr and Pinterest can be analysed.
The software identifies products ‘in the wild’, allowing brands to see images of their products shared on social media that aren’t tagged with identifying text.
It can also identify brands regardless of whether logos are partially obscured, upside down or if the image has a filter. It also analyses facial expressions and scenery to give users a fuller picture of the people who are using their products.
Ditto Labs chief executive David Rose said that the software is being used by marketing agencies and global brands in fashion, technology and automotive industries, giving its clients insight into the online presence and perception of their brands.
Retailers can use Ditto Labs’ software to “spot consumer trends earlier and understand how individual products, looks or brands are being represented”, says Rose.
By gaining insight into this “visual conversation” Rose says that retailers will be able to “adjust their marketing strategies in real-time to match or outperform their competitors”.