onton Secures $7.5M to Expand AI-Powered Shopping

AI-Powered Shopping Platform, Onton, Secures $7.5 Million in Funding

Leading technology corporations are increasingly integrating AI not only for content creation and summarization, but also to revolutionize the shopping experience. Companies such as OpenAI, Google, and Amazon are making substantial investments in AI assistants designed to aid consumers in product research and selection.

Rise of AI in Product Discovery

Several startups, including Perplexity, Daydream, and Cherry, are pioneering the use of AI specifically for product discovery. This trend has demonstrably increased customer reliance on AI during the shopping process.

Onton, formerly known as Deft, an AI-driven furniture shopping platform, reports significant growth. Its user base has expanded from 50,000 monthly active users to over 2 million, processing millions of searches and image generations.

Funding and Expansion Plans

This growth has spurred the company to announce a new funding round of $7.5 million, led by Footwork, with contributions from Liquid 2, Parable Ventures, and 43. This investment brings Onton’s total funding to around $10 million.

The company intends to utilize these funds to broaden its product categories, initially focusing on apparel and eventually extending into consumer electronics.

Earlier this year, the company underwent a rebranding from Deft to Onton, a decision made to address confusion surrounding the original name and to facilitate the acquisition of a more advantageous domain.

Addressing E-commerce Challenges with Neuro-Symbolic Architecture

Zach Hudson, co-founder of Onton, explains that while large language models (LLMs) excel at predicting user intent, they haven’t fully resolved key challenges within e-commerce.

He further notes that the time consumers spend making purchasing decisions has been increasing.

Onton’s core technology is built upon a neuro-symbolic architecture. This approach is designed to mitigate the hallucination issues common in LLMs, delivering more accurate and logical search results.

The startup’s model is also capable of incorporating real-world knowledge that may not be explicitly stated in product descriptions.

Understanding Consumer Needs

“For example, if a customer seeks pet-friendly furniture, our system recognizes that polyester fabrics offer greater resistance to stains and scratches, making them a suitable choice,” Hudson clarifies. “The tools continuously learn from each search, enhancing their intelligence at an accelerated pace.”

The AI model also accounts for variations in product naming across different websites, ensuring more comprehensive and relevant results.

Innovative Input Methods and Features

Onton provides diverse input methods to support both short-term and long-term decision-making. Users can upload images or provide prompts to visualize their desired home or office setup, and the platform will suggest appropriate furniture.

An infinite canvas with image generation allows users to combine existing images with product suggestions for ideation. Furthermore, users can upload images of their rooms and request the tool to furnish them.

The company believes these features offer consumers greater flexibility in finding what they need, even when they struggle to articulate their preferences precisely, moving beyond a simple chat-based interface.

Improved Conversion Rates

Onton reports that its approach has resulted in customer conversion rates 3-5 times higher than those of traditional e-commerce sites, attributed to the trustworthiness of the underlying data.

Expansion into Apparel and Team Growth

Hudson indicates that the technological and interface improvements implemented will facilitate the launch of an apparel category. The company is currently building its catalog for this vertical and anticipates a near-future launch.

Competition in the apparel sector will include companies such as Daydream, Aesthetic, and Style.ai.

Onton has expanded its team from three full-time employees in 2023 to ten, with plans to reach fifteen by recruiting engineers and researchers.