With almost two billion images now uploaded across the internet daily, visual content is king. But brand owners, eager to tap into image-based user-generated content (UGC) to better connect with their audiences, are struggling. For though they may be a potential gold mine, online images are, in many cases, a data blind spot.
It's hard to argue with the power of the visual online. The ubiquity of the smartphone, and now the rise and rise of image-based social media such as Pinterest and Instagram -- on which more than 40 billion images have been posted since launch -- has revolutionized the internet, which is now evolving into what some call the "visual web." Countless eye tracking studies, meanwhile, show that internet users now focus more energy and attention on visuals than anything else.
However, when bombarded with images, consumers quickly ignore those without relevance or impact, prioritizing their attention instead to visual content that is authentic and not staged. Which is why a growing number of brand owners now want to make use of real rather than commissioned images so they can engage more deeply with their audiences and better connect.
But brand owners have a problem. An estimated 80 percent of images relevant to a brand don't have relevant accompanying text, which makes them impossible to track using traditional social listening tools. They can't find the images they could use and they can't take advantage of the global photo sharing phenomenon at scale. This is why 84 percent of U.S. marketers think there is a need for advancements in image-recognition technology to help assess the context of an image without text, according to a recent survey.
Help is at hand, however, and from an unexpected source: software engineers that build sophisticated computer vision programmes that can track and identify millions of different images online recognising everything from a person's facial features to everyday objects, locations, and even human emotions. Current estimates suggest the global image recognition market will be worth $33.3 billion by 2019.
Already, "in-image" advertising exists that allows advertisers to automatically place relevant digital ads within editorial images using image recognition. Crucial to this is relevancy. Technology can detect, for example, an image of a person running and insert an ad from a sports manufacturer.
In-image advertising's power lies in its ability to blend in without detracting from the content the consumer is there to see. And it is already delivering results with viewability rates of up to 80 percent more than that achieved by traditional digital ads
The work of "image scientists" is opening up new opportunities for content publishers to monetise editorial images on their platforms for the very first time. And they are offering brand owners premium inventory they've never before been able to access.
Looking ahead, as the visual web evolves further -- as it surely will -- and images increasingly replace words, the role "image science" plays in marketing will only grow, with software making it possible for brands not only to analyze what people are writing about them, but to see and understand how consumers truly feel about them.
Reading between the lines -- and delving into what hasn't been written next to an image, but what is actually in that image -- will provide for brand owners new and rich data about their consumers. And for a brand owner in today's highly competitive marketplace, surely, this depth of nuanced, actionable insight will be as good as any proverbial pot of gold.