Twiggle, a Tel Aviv-headquartered startup focused on making searching for products online as simple as talking to a store clerk, announced today that it has raised $15 million led by MizMaa Ventures and Korea Investment Partners.
Konigsberg told TechCrunch in an email that Twiggle will use its new capital to scale up its Semantic API, which allows companies to use its technology without replacing their existing search engines, add new search-related features, hire more executives, and grow its U.S.-based sales team.
Twiggle’s natural language processing strives to replicate the experience of walking into a store and asking a sales associate to help you find something by supporting complex queries. Instead of trying to brainstorm keywords, for example, customers can simply enter conversational sentences describing what they want to order.
The company hasn’t disclosed which e-commerce sites use its technology (Konigsberg says only that it’s “some of the biggest retailers in North America, Asia, and Europe”), but they presumably include platforms owned by investors Alibaba and Yahoo Japan, as well as Naspers portfolio companies (which include Flipkart, Avenida, and OLX).
Having Mizmaa Ventures (an investment firm that focuses on bringing technology from Israeli startups to China and the U.S.) and Korea Investment Partners will help Twiggle’s expansion plans in Asia.
“The addition of MizMaa Ventures will strengthen our ties with the Chinese retail market, and Korea Investment Partners has opened up new opportunities in Korea, with leading e-commerce players such as Coupang,” says Konigsberg. (Coupang has been described as the Amazon of Korea).
The ability to layer Twiggle’s search technology onto existing search engines is one of the startup’s main advantages against Amazon’s A9, which is also available to third-party retailers.
“Search is core to our customers’ businesses and they want to be able to control and customize it based on their own business needs. But they’re also always looking to leverage the latest technologies to improve performance and gain a competitive advantage,” said Konigsberg. “That is where we can help.”
Twiggle plans to add feature extraction from product images, in addition to textual content such as product descriptions and reviews, in order to improve its search results and product recommendations. Konigsberg claims that Twiggle has increased relevance and clickthrough rates on search results by up to 30 percent and conversion rates by around 20 percent.