Originally published here by ClickZ
Green shoots are tentatively emerging for retail, with the latest Barclaycard data showing an upturn in spend. And with the pandemic driving consumers to increase their digital shopping – online retailers currently claim £3 in every £10 spent – it’s vital retailers rethink their engagement strategies to deliver immersive online experiences that really drive conversions.
Shoppable video that allows consumers to interact with and buy featured products is proving a popular solution, particularly when enhanced with artificial intelligence (AI) and computer vision, which significantly increases consumer engagement and conversion rates. So what exactly is computer vision and how can it be used to boost the performance of shoppable video?
Computer vision is a long-established segment of AI which deals with a machine’s ability to understand and process visuals and provide appropriate outputs. It is already widely used in a variety of applications such as facial recognition software and photo tagging features on social media platforms. But as computer vision is developed and refined, it is increasingly being used in retail marketing to understand consumer needs and streamline the purchase process through channels such as shoppable video. Here are three ways computer vision and AI can bring shoppable video to life:
Real-time product recognition
Computer vision enables machines to look at visual media such as images and video, and extract detailed attributes such as style, colour, and pattern in the same way as the human brain. These attributes can be used to identify retail products featured within visual media in real time and connect the consumer with those products.
Brands and retailers can upload entire catalogues and use computer vision to map all products contained within them. These products can be automatically recognised within video content without retailers having to spend hours tagging or assigning them manually. Shoppable hotspots can then be created that give the consumer an immediately satisfying experience where they can go straight from product interest to purchase, all without leaving the video environment.
As well as identifying individual products in real time, computer vision can be used to recommend similar products, or those that complement items the viewer is interested in. By detecting keyframes within video content and automatically identifying featured products, the technology can instantly generate alternative suggestions and recommendations from the retailers product catalogue.
Shoppable video with computer vision enables the consumer to open and browse curated product selections without navigating away from the video or being redirected through multiple pages and links. For instance, if a consumer is viewing a promotional video for a homeware brand’s new range of garden furniture, a shoppable hotspot can encourage them to browse and buy complementary accessories such as cushions and lighting, as well as chairs and tables from the furniture range itself.
While suggestions for alternative or complementary products are predominantly based on products within the video content the consumer is viewing, an additional layer of personalisation can also be added into the equation. Recommendations can take into account data relating to the user’s demographics, preferences, and past browsing or purchase behaviour to ensure the products displayed are those most likely to appeal to their individual tastes.
Analytics and insight
AI enables deep and ongoing evaluation of visual content to help brands and retailers refine and optimise their video strategies. It allows them to move beyond generic metrics such as view rate to fully understand viewer experiences. They can determine at which point in video content viewers are most engaged, which hotspots are interacted with most often, and where consumers lose interest and drop off. They can understand highly specific preferences such as favoured music tempo, as well as analysing unique attributes such as device type and geographic location.
By feeding these insights back into content creation, brands and retailers can continually improve the performance of shoppable video. When they understand what is working and what isn’t they can make real-time adjustments to current content such as fine-tuning video imagery, audio and duration. They can also use the insight to inform and refine future video experiences and ultimately achieve better return on investment (ROI).
As the shift to online shopping accelerates, brands and retailers should look to shoppable video to provide immersive and interactive digital experiences. When it is enhanced with AI and computer vision to enable real-time product recognition, personalised product recommendations, and actionable analytics, shoppable video delivers exceptional engagement and conversion rates, allowing brands to make the most of any upturn in consumer spend.