During the pandemic, retailers were forced to embrace e-commerce. But some have found they struggle to maintain customer loyalty as consumer needs change and shopping patterns change. With fierce competition from the likes of Amazon, customers have proven to have little patience for websites that don’t deliver what they want. According to Baymard Institute research, for every 100 potential customers, 70 leave without buying.
That’s why Purva Gupta started Lily AI, one An AI-powered platform that connects retailers or brand consumers with the products they want to buy. Co-founded by Sowmiya Narayanan, Lilly provides algorithms designed to power web store components such as search engines and product discovery carousels.
Lilly today announced it has raised $25 million in Series B funding led by Canaan, bringing its total funding to $41 million.
“Different consumers are looking for something unique, making it important for retail ecommerce brands to create common, long-tail searches to build the right product taxonomy,” Gupta told TechCrunch in an email. “About disappointing experiences on retail ecommerce sites and irrelevant results or worse, no results at all, even though the retailer clearly carries the product you want.”
Prior to co-founding Lilly, Gupta held various roles at Eco India and UNICEF. Narayanan at Texas Instruments Yahoo! (Full disclosure: TechCrunch’s parent company) and Box, she was a full-stack web dev for the product Box Notes.
Lilly started life as an app for retailers to understand the personal preferences of women shoppers around fashion. But when traction is hard to come by, Gupta and It inspired Narayanan to build a more enterprise-focused solution packaged as a plug-in, software-as-a-service subscription product.
Lilly now maintains a team of fashion, home and beauty “experts” who help refine product taxonomies, which are then used to train algorithms for product discovery and recommendations. (The team researched and developed ways to convert things like “ribbed fabric” and “minimal clothing style” into a mathematical “language” so the algorithms could help.) Basically, Lily stores detailed information on products based on attributes (e.g., “style,” “fit,” and “coincidence”) and uses customer data derived from product names associated with the item’s attribute data to predict each customer’s affinity for product attributes in the catalog.
Gupta acknowledged that there are other companies in the product labeling and automated product labeling areas that rely on automation and AI. For example, Depict.ai offers a product recommendation tool that pulls from data from across the internet. Black Crow AI is developing a platform to predict which products e-commerce customers will buy, while Constructor sells a framework that powers the search and discovery of digital retail markets.
Meta has also tested clothing behavior prediction on the Facebook marketplace, showing a system that can extract clothing characteristics and fashion styles from photos of models on Instagram and Flickr two years ago.
But she argues that Lily is one of the most powerful options in setting up. Gupta also emphasized that the platform is as privacy-friendly as possible, not using customer names, addresses, or financial transaction information to anonymize user interactions on its customers’ ecommerce sites.
“The IT decision makers we work with are focused on a more realistic and realistic application of Lilly and being on the strategic front. They are interested in the depth and accuracy of the information that Lilly can provide; how we train the models; and the accuracy of the results and confidence levels,” she said. “We win by customizing our product to meet their needs and have a dedicated customer success team available to take into account changes in goals or outcomes over time.”
Regardless, big-name clients have signed up for Lilly’s services to date, including Macy’s, The Gap and its various brands, Bloomingdale’s and thredUP.
Lilly is loathe to make its revenue figures public, and the 87-employee company said it did not have a headcount forecast for the end of the year. Gupta brushed aside questions about the volatility, saying Lilly was “well positioned” to take advantage of new retail assets in the coming months, even if macroeconomic headwinds were to hit.
“Lily AI has grown exponentially. Since the start of the pandemic, health crises have accelerated retail’s shift to e-commerce and digital transformation,” Gupta said. “We will use the new funding to further expand into enterprise and mid-market retail e-commerce products in home, beauty and fashion… We also plan to further extend our solution to additional applications in the retail stack. A rich set of analytics for our clients.