Visual content is a primary canvas for marketers to tell stories that influence their consumers. Videos, images, and everything between (and beyond) dominate websites, social platforms, and apps, providing marketers with innumerable ways to bring their messages and stories to life.
However, with images and videos populating all corners of the digital landscape, it is challenging for them to ensure their marketing campaigns stand out or reach the right people at the right time. This visual clutter has literally left marketers poking around in the dark, struggling to get their targets or derive meaningful insights from their campaigns.
Thankfully, in recent years, several new tools have emerged to help marketers take control of their visual strategy and improve the performance of marketing efforts. One of those is computer vision (CV), a field of artificial intelligence that trains computers to interpret and understand the visual world.
What is computer vision? Computer vision is a technology typically applying the technique of “deep learning” that allows machines to identify and classify objects or data within an unstructured format more accurately than ever before. It seeks to automate classification tasks that the human visual system performs and trigger appropriate actions based on that classification.
Key applications of computer vision
- Object Classification – Classifying different objects and faces into various categories
- Object Verification – Checking if an image matches its identity.
- Object Identification – Identifying which type of a given object is in the image.
- Object Recognition – Recognizing what objects are in the image and where they are.
- Object Detection – Detecting wherein an image is a specific object.
- Object Tracking – Tracking a specific object across a series of images
- Semantic/Instance Segmentation – Breaking down the object into its components, including counting
- Facial Recognition – Identifying gender, age, cultural appearance, emotions, etc.
- Action Recognition – Identifying a specific action/gesture of a person
- Mood and Sentiment – Forecasting someone’s reactions or current mood
- Crowd Dynamics – Counting people and tracking their density/direction
- Object Character Recognition (OCR) – Recognizing what is written in a particular image (i.e., text and numbers)
- Document Analysis – Analysing a document and providing the information enquired about
Computer vision in marketing
Computer vision gives machines the ability to “see” imagery through mathematical representations of three-dimensional shapes and appearances. It allows computers to identify subject matter in an image and comprehend meaning and context similar to humans. This opens up a wide array of potential uses for marketers.
1. Image recognition and content discovery
Image recognition is one of the key use cases of computer vision in marketing. Using machine learning algorithms and pattern recognition to acquire human visual systems’ veridicality, computer vision enables them to understand the digital images and their content and enhance the customer experience. Recent improvements in image recognition are incredibly useful tools for marketers to implement, especially given consumer’s preference for visuals in digital marketing efforts. Image recognition and computer vision can vastly improve how marketers understand, track, and interact with and understand consumers at scale in a way that works congruently with their lives.
2. Segmentation and targeting
Computer vision offers entirely new ways for marketers to segment and target customers. It allows marketers to achieve a deeper understanding of their customers and segment accordingly. Usually, the pictures that individual shares on various social media channels reveal valuable insights into what makes them tick. This information can then be used to segment customers to be targeted with personalized advertisements. By analyzing customer behavior in an organic format, marketers can create increasingly accurate customer segments with a much higher chance of a marketing message resonating with them.
Some companies are already implementing this into their marketing efforts. For example, Coca-Cola’s Gold Peak iced tea brand utilized image recognition technology to scrub through Facebook and Instagram, then find people who are drinking iced tea and exhibit happy emotions. These indicators in their pictures served as self-assigned customer segments that Gold Peak then used to target ads. Once users left their social media platform, they were targeted with Gold Peak ads on mobile and desktop websites. This is just one example of how computer vision can be used to segment and target customers intelligently.
3. Social media engagement
Marketing in today’s digital age is a dynamic relationship between the marketer and consumer. The billions of people who use social media possess the powerful ability to co-create and interact with brands through viral content, user-generated content (UGC), influencers, and facilitating word-of-mouth (WOM). Marketers’ ability to understand and analyze these organic interactions’ impact can be crucial to their overall performance.
Marketers can measure their social campaigns’ effectiveness through tracking engagement rates, impression counts, link clicks, and more. Still, these metrics are limited to basic ways of analyzing customer behavior and understanding sentiment accurately. Traditional social media metrics are limited to tracking specific behaviors that do not always provide clear insights and may exclude people who do not like participating in those behaviors (e.g., a person who does not like sharing posts from companies on their social media account).
Computer vision can significantly impact social media marketing by tracking and analyzing behaviors that provide informed insights to marketers. Consumers reveal incredibly valuable information about themselves through the photos they post, share, and engage with. There previously had been no way for marketers to tap into this potential at scale. By using image recognition and computer vision applications, marketers can identify which brands consumers are posting about, how they use them in their daily lives, their role in candid interactions, and much more.
4. Interactive marketing
Interactive marketing is one of the top use cases of facial recognition in marketing today. Interactive marketing operates by reading the participant’s biometrics, allowing the brand to capture data on their sentiment and mood while offering a fun experience. One exciting example of this is Expedia’s Discover Your Aloha campaign. Expedia’s campaign works by having the customer turn on their webcam, allowing them to be immersed in a tropical paradise as they are led through Hawaii’s beautiful scenery.
Where facial recognition technology comes into play is by analyzing which part of the experience elicited the visitor’s most positive reaction. Once visitors left the site, they were targeted with a coupon for the area in Hawaii, that they responded more positively. This creative use of facial recognition allows Expedia to create an interactive, experiential marketing campaign to delight its customers and add more value by offering highly personalized discounts to let visitors see their favorite part of Hawaii in real life.
