It is impossible to open a social network these days and not see the stream of images, videos, and graphics that dominate the experience.
In fact, according to a survey by Harvard University, Snapchat is the fastest growing platform among Millennials, with Instagram close behind. And while Facebook still reigns king as the most widespread social network, with 81% of survey respondents owning an account, there’s no denying that video and imagery have dominated Facebook updates in the last year or two.
So what does this mean for brands? How can companies and marketers capitalize on the opportunities presented by the increasingly visual Web?
The Value Of Image Monitoring
In 2014, 1.8 billion photos were shared every day on social media. Do the math: that’s over 1 million images a minute.
For businesses of all kinds, this is a good thing: brands now have the ability to incorporate visual content into the list of ways they can better understand their customers, engage around topics of interest and affinity, and know how their brand is being portrayed across social media channels.
Are the most popular images the ones we want associated with our brand? How is our content performing vs the competition, and what of our best-performing content should we consider amplifying with paid media? Is our own content performing best, or is content created by our community capturing more attention and interaction?
Beyond reputation management, visual content can also provide examples of how products are being used and adopted in the market. For example; popular “hack” images and videos could give rise to new and expanded opportunities, or even expose risks or concerns that a company might want to address immediately. For businesses who have prominent (and frequently imitated) brands or logos, monitoring for copyright infringement or fraud might be important in making sure counterfeit products don’t enter the market.
But doing all this means that our existing “listening” technologies must advance beyond keyword and text-based searches, and delve into the world of finding relevant pictures and videos in the vast and ever-growing sea of visual content.
Technologies To Watch
The challenge for brands to date is that listening and monitoring – cornerstones of modern digital brand management – have been limited to text analytics alone. Unless images were accompanied by text / a caption, there was virtually no way of finding them (and not all technologies were even able to capture that text to start with, depending on their data input and capture capabilities).
Today, solutions that blend the power of text analytics with powerful image-matching and recognition algorithms are emerging. They can identify objects – like a shoe, or a golf club, or a hamburger – within an image, and find a distinctive logo or graphic within a photo no matter its size, orientation, or color. If you’re a soft drink company, for example, and want to find professional golfers enjoying your beverage on the course, that’s pretty powerful indeed.
Just last year, Sysomos added GazeMetrix to our platform for exactly this reason. Combining deep learning technologies – a type of abstract machine learning wherein the technology can effectively “remember” what it sees – with text analytics and powerful image-matching algorithms, Gaze brings these capabilities to life.
And that’s just the start. Today we can do amazing things with static-image recognition. Later this year we’ll extend these capabilities to videos as well.
How This Might Work
Let’s look at a real-life example to illustrate how this might work:
A beer manufacturer had been largely targeting their advertising and marketing to a leisure crowd, with ad spots featuring people enjoying their beers on a beach / vacation.
Once they added image monitoring to their capabilities, something very interesting came to light: a huge portion of the images that people shared (ones featuring their brand of beer) were not on beaches or vacation, but rather right at home during their backyard barbecues with friends.
Leveraging what they learned, the company shifted some of their advertising visuals and messaging to better reflect the real-life enjoyment of their product, increasing the effectiveness of their advertising dollars amongst their target audience.
Where Do We Go From Here?
The visual content space is certainly not perfect yet, and image recognition technologies are really just starting to evolve. But with continued advancements in artificial intelligence, in machine and deep learning capabilities, and in available data sets – think Internet of Things and wearable data – we will soon be able to move further, faster.
The truth is: we’re only at the start of a critically important innovation in advanced analytics ; one guaranteed to change the way companies “listen” to their customers.
There’s no denying the power of visual content and storytelling; it’s critical to brands already. But the more we support the ability to analyze this type of content, the better we can help marketers and consumers alike harness photos and videos to drive an immersive, personalized experience.
Over the next five years, it will be amazing to watch our industry evolve from simply listening to seeing, hearing, and immersing ourselves in the customer experience through the power of visual content.