Visual AI improving online shopping experience
The poor product discovery experience of consumers is costing online sellers and e-commerce companies billions in lost sales as they fail to live up to user expectations. Helping them address this problem, visual AI startup for e-commerce, Pixyle AI, has raised a €1M seed funding round from South Central Ventures. Their mission is to connect retailer’s online audience with the products they are really looking to buy.
Founded in 2018 by Svetlana Kordumova, a PhD in AI and Computer vision from the University of Amsterdam, Pixyle AI helps fashion retailers with image recognition solutions. It is enabling them to deliver a better online shopping experience, boost e-commerce sales, and improve efficiency in catalogue management with smart product data entry.
Pixyle’s visual AI finds the location of fashion items in images and determines the category of the item i.e dress, sneaker etc. Then Pixyle AI digs deeper into the image to classify attributes, i.e. colour, pattern etc in language that shoppers use. This enhances the product data and improves the relevancy in search results for product discoverability, leading to higher conversion rates. The tags generated by Pixyle AI can be easily absorbed into a retailer’s existing e-commerce stack, making their search and discovery platform run seamlessly and faster.
Svetlana Kordumova, founder and CEO of Pixyle AI commented: “In search, understanding customers’ shopping intent is vital. We are laser focused on this using neural networks that train Pixyle’s AI algorithms to see and interpret images exactly how humans would. For example, in e-commerce when site visitors are looking for a specific product to buy and type “short summer dress with flower print in purple and pink” in order for these kinds of dresses to be shown to the customers, in the catalogue they need to be tagged with length “short”, occasion “summer”, with “floral” pattern, and “purple” and “pink” colours. It starts with building the right taxonomy. Having the right metadata associated with product images helps retailers capture everyday popular search queries as well as long-tail searches. We’re trying to make online shopping as frictionless as possible, and hopefully a bit more fun and enjoyable by letting people easily find what they are really looking for.”
Increased relevance of the search results
The premise behind Pixyle AI is that search engines and site search are failing shoppers. The inconsistent, incomplete and inaccurate data quality linked to products isn’t helping them find what they are looking for. Research shows that 94% of searches deliver irrelevant results and 72% of sites completely fall short of search expectations. As a result, the majority of shoppers abandon their search and they’re not coming back. Google has estimated (in the US alone) the cost of such bad search experiences is $300B in lost revenue each year.
Jan Kobler, Managing Partner at South Central Ventures commented: “E-commerce has gone from strength to strength in recent years with several innovations to help the channel become second nature for consumers. A pivotal part of engaging online shoppers is product search, being able to find what you want easily and quickly. However, search has been hugely underserved and remains an unmet need for retailers and shoppers until now. Pixyle AI is laser focused on this opportunity and is already moving the dial with more sales for retailers. They have built a robust tech stack, which has been tried and tested in the market and is ready to scale. We are very much looking forward to working together with the founders and helping them grow their business in years to come.”
Svetlana Kordumova added: “Retail footfall has increased post pandemic but online shopping for clothes will continue to grow. We are also encouraged by the growth of the fast-growing fashion resale market, which presents a significant first-mover opportunity for Pixyle AI to help the circular economy take off. We have built algorithms specifically for user-generated images, allowing the AI to detect and recognize fashion items even in blurry, homemade selfies.”
About Pixyle AI
Pixyle AI, with offices in Amsterdam and North-Macedonia, helps brands and retailers all over the world. Their clients include Tendam (Spain), Otrium (Netherlands) and Miinto (Denmark). Since going live in 2019, Pixyle AI has worked with retailers having tagged over 250M images. Their algorithms are improving the relevance and accuracy of product attributes with 95% accuracy and increase conversions by 10% on average. The seed funding will allow Pixyle AI to further expand its geographic reach, enhance the product offering, and move into new verticals.