By Ben Plomion, GumGum
By now users of apps and devices like Siri, Cortana and Amazon Echo are accustomed to the idea that machines have gotten really good at understanding human language. Voice-recognition technology has changed the game for millions of consumers — and for marketers. Amazon, for instance, is thrilled that consumers can now bark orders (especially merchandise orders) at the Echo’s virtual assistant, Alexa.
But there’s another parallel phenomenon going on: Machines have also gotten really good at understanding the images that humans create. This great leap forward couldn’t be coming at a better time, given the ubiquity of devices and platforms that allow us to effortlessly create and share images, and the resulting explosion in our visual culture. Facebook alone says that users upload more than 350 million photos to its servers each day.
Advertisement
Who can keep track of all those images?
Machines can — or, more accurately, machines equipped with image-recognition software that can “look” at images and make sense of what’s in them.
As image-recognition software grows ever more powerful thanks to artificial intelligence deep learning techniques, consider the ways in which the marketing world will be transformed:
Visual Trendspotting
Nearly 20 years ago in a New Yorker article titled “The Coolhunt,” Malcolm Gladwell wrote of a new class of market researchers he dubbed “coolhunters.” It’s hard to imagine now, but this was in a dial-up world where Google didn’t yet exist.
Social media helped make the art and science of trendspotting easier — or at least more comprehensive and widespread. Plenty of marketers now use social listening tools to identify market trends, but those tools remain, for now, largely text-centric: They tune into what consumers are saying.
But with more and more consumers using social media to show rather than tell (think of the Millennials who live on Instagram and Snapchat), image recognition has the potential to deliver unprecedented insights into what’s actually happening in the marketplace by tuning into what consumers are showing in the pictures they share.
In the fashion world, for instance, image recognition will be able to tell us what styles and brands of clothing people are really wearing in specific cities and across regions. Companies in the design space like Pantone — which each year announces a “color of the year” in advance to great fanfare — will have the ability to detect what hues are actually rising and falling in fashion and products, in real time, across millions of up-to-the-minute images.
Evaluating Products In Use
Image-recognition software is growing better and better at product recognition, thanks to visual cues like distinct shapes and logos that can be algorithmically codified. It’s also getting ever-better at deciphering context.
If you’re a marketer, you want to know exactly how and when consumers are using your product. When a suburbanite buys a 4×4, for instance, is it actually for frequent offroading, or is it more of a lifestyle statement? Image recognition will be able to tell us, no doubt because that driver will be sharing images containing that car on social media.
Gauging Consumer Satisfaction
Every consumer who has a late-model smartphone is carrying around a computer equipped with basic image-recognition software. When you point your phone at your friends to take their picture and the built-in camera squares off their heads so as to automatically set the focus on their faces, that’s image recognition at work.
More powerful computers running more powerful software — like the machines and algorithms that are part of Microsoft’s Project Oxford — can not only identify faces but detect the emotions being expressed by those faces, including happiness, sadness, anger and contempt.
Imagine a world in which marketers can supplement — or even bypass — focus groups, surveys and other market-research tools in favor of image-recognition technology that can make sense of both products in use and the mood of the consumers using those products.
It’s coming.
These days marketers take for granted that social is the essential new arena for market research. But as visual social continues to take off — and next-generation image recognition makes sense of it — the world of market intel is poised for a whole new revolution.
Ben Plomion is the SVP of Marketing at GumGum and brings more than 15 years of experience in marketing, communications and also business development. Prior to GumGum, he was responsible for Chango’s brand, integrated marketing and demand generation. His team created one of the most robust thought leadership platforms in the industry and has won multiple marketing and design awards. Prior to joining Chango, Plomion worked with GE Capital for four years to establish and lead the digital media practice.
