NEW YORK: If a picture is worth a thousand words, images at scale-to the tune of several billion-are a potential treasure trove of data.
Brandwatch launched an image-recognition platform, called Brandwatch Image Insights, on Wednesday to help brands track and analyze logos.
"Images are the world’s common language," said Phillip Agnew, product marketing manager at Brandwatch. "[About] 3.2 billion images are shared online daily. By 2020, that’ll be 10 billion. And that’s a hell of a lot of data. At Brandwatch, we want to be the company that understands images at scale."
The platform currently pulls images from Twitter, with plans to grab them from Instagram in a few months. It could help brands determine the performance of advertisements, identify what influencers are interacting with their products and services, and discover user-generated content. Brands can also figure out how their logos and imagery are being misused.
"Social is the world’s largest focus group," Agnew said. "With images, you get even richer input."
Agnew explained the platform is built on machine-learning-enabled analytics and detection capabilities that present multiple case uses. Brandwatch developed its image analytics and logo detection features for 18 months.
"We can cut up the logo into smaller pieces and feed the technology those pieces, so it can learn to find even parts of a logo in an image," he said. "[The AI] will learn every time it’s used."
Agnew noted that the simpler the logo or image, the more difficult it is to track. He claimed Brandwatch’s platform can keep tabs on logos as simple as Nike’s swoosh.
Brighton, U.K.-headquartered Brandwatch has U.S. offices in New York and San Francisco. Its platforms are used by companies including American Airlines, Walmart, and Dell, according to its website.