Syte raises $40 million to bring visual search to online retailers

Visual search is now a core part of countless mobile apps, from Google Lens and Bing to Pinterest and eBay. But building the underlying AI infrastructure to let people search databases and catalogs using images rather than words is a resource-intensive endeavor requiring a lot of time and technical expertise, which is where Syte comes into play.

The Israeli startup has developed what it calls “the world’s first product discovery platform,” allowing companies of any size to integrate visual search into their own software via Syte’s API. Today, Syte announced that it has raised $30 million in a series C round of funding, plus an additional $10 million in debt financing, to expedite its global expansion and further develop its visual discovery engine.

Founded out of Tel Aviv in 2015, Syte has developed the visual AI smarts that retailers can use to improve the search experience in their online stores. For example, a fashion brand might want to allow customers to search for new items using a photograph that they upload, or find similar items to something that they see on the brand’s website or social media channels.

Conversely, a retailer might want to incorporate a recommendation engine, whereby it proactively suggests different types of clothes based on something that a customer has proactively searched for — and this is something that Syte offers too.

To integrate with Syte initially, retailers first have to send their entire product catalog to be indexed and tagged. After that, Syte continuously monitors and indexes the customer’s product feed throughout the day to ensure that it’s up-to-date.

“This way, shoppers are presented with the most relevant results in real-time and brands can provide hyper-personalized experiences based on shoppers’ visual preferences,” Syte cofounder and CEO Ofer Fryman told VentureBeat. “Additionally, the automatic product tagging enables brands to easily set up smart merchandising rules to, for example, promote a seasonal collection or best-selling items within search results.”

The state of play

There are other visual search providers out there, including Microsoft’s Bing Visual Search and VC-backed New York startup Donde. However, Syte claims that its speed and 95% accuracy in object matching sets it apart, while visual search is only one component of its offering. For example, earlier this month Syte launched what it calls “augmented site search,” which uses visual AI to automatically categorize products using the type of language that a shopper might actually use in natural language searches, such as “red dressy heels.”

Above: Syte: Augmented site search

“Because Syte’s visual AI is consistently trained on fashion, jewelry, and home decor, including millions of products from vendors all around the globe, it has amassed the largest vertical-specific lexicon in the industry, enabling it to not just recognize objects within an image but also to assign them a variety of detailed product tags based on their visual attributes,” Fryman said.

This has the potential to greatly improve the accuracy of search results on ecommerce websites. For example, a traditional keyword search for a “dress shirt” would likely surface items of clothing that have the words “dress” or “shirt” in their descriptions. But when an item is automatically categorized based on its visual appearance, this can help reduce the number of steps that a shopper needs to take to find what they’re looking for.

Above: Syte: A traditional ecommerce search engine (left) versus augmented site search (right)

Prior to now, Syte had raised around $31 million, including its $21.5 million series B round last year, and its latest funding round was led by return investor Viola Ventures, with participation from LG Tech Ventures, La Maison, MizMaa Ventures, Kreos Capital, Magma, Naver Corporation, Commerce Ventures, Storm Ventures, Axess Ventures, Remagine Media Ventures, and KDC Media Fund.

With another $40 million in the bank, Syte will expedite the launch of what it calls a “complete product discovery” platform. “This will allow retailers to create, personalize, monitor, test, and analyze all of their websites’ discovery journeys in one place,” Fryman said.

Additionally, Fryman said that the funding will help scale Syte’s R&D efforts and expand its augmented site search to more languages, while it may also explore acquisitions where necessary.