E-commerce has been growing exponentially in India and around the world. But can visual search and social commerce take it to another level and make the end-user experience better?
While e-commerce has brought about a paradigm shift in the way people shop for products, from browsing at offline bazaars to scrolling on online marketplaces, consumers are now overburdened with sifting through large amounts of products on different e-commerce platforms. Finding the specific product one is looking for is a complex process and could be as challenging as searching for a needle in a haystack. One needs to search with the right keywords and hope that products are correctly tagged and categorised on different e-commerce platforms.
Visual search and social commerce seem to be the answers to this pain point and some experts even believe that they could be the frontiers on which the future growth of e-commerce will depend. YourStory recently interacted with Oliver Tan, CEO of ViSenze, an artificial intelligence (AI) startup working in the e-commerce space, to better understand the visual commerce space, the potential it has, and the work ViSenze has done.
Story so far
ViSenze was founded in 2012 by Roger Yuen, Oliver Tan, Chua Tat-Seng, and Li Guangda. Talking to YourStory, Oliver recalled that ViSenze was literally born out of a lab at the National University of Singapore when he and his founders realised the potential impact it could have on e-commerce.
With a bachelor’s degree (honours) in economics from the University of London and an MBA from Royal Holloway, University of London, Oliver has over 19 years of diversified business experience in the technology, media, and telecommunications (TMT) space. An former venture capitalist himself, he noted that he enjoys working with startups, having spent five years as a key staff in a technology startup till its successful exit.
ViSenze has so far raised about $14 million in funding from firms like Rakuten and WI Harper. A 40-member team, ViSenze consists of web specialists and computer scientists with deep machine learning and computer vision experience. It has offices in the US, UK, India, and China and is headquartered in Singapore.
The startup powers visual commerce at scale for retailers and publishers. It delivers intelligent image and video recognition solutions that shorten the path to action as consumers search and discover products on the visual web. Describing their journey over the last few years, Oliver said,
"We have transitioned from machine learning to deep machine learning and work on fundamentally convoluted neural networks. In simple terms, we teach machines to use different layers of logic and try to guess and make decisions as humans would."
Visual commerce leveraging pixels
Oliver informed that ViSenze has built technology that allows them to process pixels and gain intelligence from images. He believes that this sets them apart from other visual search startups, many of whom are focused on processing metadata and not on images. Talking about the pain point, Oliver said,
"Ninety percent of images on the web cannot be found simply because they aren’t tagged or not tagged well enough. Why are people continuing to search the web using meta-keywords? Why can’t we process pixels and images and use this intelligence to power visual search queries?"
He noted that people interact with a lot of products in their day-to-day lives and generally seek to buy them. But they struggle to describe the products in words and recognise them on the web to complete their purchase. Retailers like Rakuten and ASOS use ViSenze to convert images into immediate product search opportunities, improving conversion rates. Oliver recalled that while Rakuten is one of ViZense's lead investors, they were, in fact, quite strict while evaluating their portfolio company's technology and signing up as a customer.
At this stage, Oliver took out his smartphone and gave me an overview of how their technology works. He showed me a pair of heels with straps and we tried searching for them on e-commerce platforms with different keywords. The results were not very accurate. He then used ViSenze’s APIs to search for a similar product on Amazon and we found an accurate match immediately. Oliver said,
"Pictures may be worth a thousand words. But do we really need a thousand words to describe it?"
So ViSenze works with different horizontal and vertical e-commerce players across the world and in India to add a visual search experience into e-commerce. Some of the startup’s other customers include Myntra, BlueStone, CaratLane, Goodrich Global, Zalora, and FlashFrames by Essilor.
Oliver said that with their visual search technology they can apply different search filters related to pricing, similarity, availability, and geolocation to increase relevance for users. He added, “We augment intelligence in data stack to be sharper and more relevant instead of just accurate. We help retailers optimise keywords and also augment the catalogue for best results.”
Visual search for publishers?
Oliver noted that ViSenze has the capability to process videos at scale and understand potential monetisation opportunities. Coming under the category of video commerce, ViSenze works with publishers to make subtle product recommendations in videos and help consumers purchase directly from the video. Media companies use ViSenze to turn any image or video into an engagement opportunity, driving more new and incremental dollars.
This is where social media, influencers, and celebrities play a big role. Oliver noted that online commerce is heavily geared towards recommendations and publicly admired celebrities have the biggest potential to push products, whether on virtual shelves or real. ViSenze aims to decrease the lead times during such cases and help brands and publishers cash in on the craze.
"This allows us to pull images on HTML pages and return overlay recommendations or in simple terms a ‘buy button’ when someone moves the mouse over the particular image or media."
For every successful transaction through the e-commerce platform, ViSenze shares the revenue with the publisher and brands. While this can be applied across categories, Oliver noted that fashion is currently the biggest category globally but others like eyewear, home décor, and consumer packaging products are also showing some traction.
Oliver said that on the visual commerce front they work with media companies to help them gain incremental revenues over their current ad-supported revenue models. He added,
"The road to e-commerce growth is paved with content. Retailers today are becoming publishers, why? Because they understand that content is the primary motivator and influencer for shopping behaviour."
He noted that social media platforms like Facebook, Pinterest, and Snapchat will play a key role as they look at different ways to generate or influence shopping behaviour with content. So with ViSenze, Oliver’s goal is to make every image and video shoppable.
On being asked if ad blockers could pose a challenge in this space, Oliver noted that while they are generally used to block annoying full-page popup ads, their shoppable images are non-intrusive and the ‘buy button’ is visible only by hovering a mouse over the image.
The future of e-commerce in the USA, India, and China
On being asked to share his estimates on what percentage of e-commerce will be driven by visual search in the next few years, Oliver linked it directly to digital content marketing spends. He says,
"About $145 billion is spent annually on content marketing globally, right now. This powers a $1.7 trillion e-commerce market worldwide, across different categories. This is expected to triple within the next three years."
Having closely followed markets like the USA, China, and India, Oliver noted that the first is their biggest market and is ahead in terms of adopting new-age tech tools quickly. He also noted that the e-commerce market in the USA is currently dominated by Amazon, which is ‘eating everyone’s lunch’ and players like Walmart are now doubling down on e-commerce, looking to level the playing field.
Looking at markets like India and China, Oliver noted that they have great potential to leapfrog from ‘E-commerce 1.0’ to ‘E-commerce 2.0’ thanks to the rapid adoption of smartphones and other technological innovations.
But to succeed in developing markets like India, Oliver believes that edge-case computing technologies will play an important role. In simple terms, this means that since networks are slower and heavily taxed, ViSenze aims to shift some of the processing load off the cloud and perform it on the device. This also has the added advantage of providing a smoother end-user experience.
Also unlike the USA, Oliver noted that customers in India and China are generally not very loyal to a brand. They instead focus on finding the best deals and are also influenced by celebrities. This opens up opportunities for social commerce.
Democratising AI for e-commerce?
ViSenze’s long-term vision is to democratise visual search for retailers and publishers and do the heavy lifting behind the scenes. So the company currently generates revenue through a B2B SaaS model for their services and APIs.
While many e-commerce players like Amazon and Flipkart are building visual search capabilities in-house, Oliver notes that the market is still huge with a lot of smaller vertical players across niche categories like home décor and furniture who would prefer to focus on their core business strategies and let companies like ViSenze power their visual search integrations.
Having worked in this space for five years, Oliver noted that he likes to have futuristic conversations with both his large and small customers to better understand their long-term needs, continue innovating and building, and also stay ahead of the game. He noted,
The real world is about shopping with all our five senses, so the more of these you can replicate in the online world, the better the experience you can give customers and optimise for their behaviour.
Oliver believes that voice-powered AI (like Alexa) and image-powered AI have the potential to disrupt e-commerce and influence customers even before they walk into or browse real or digital stores.