Remark puts thousands of human product experts into AI form

Shopping is fun, but making a decision on one set of skis among thousands of possibilities is not. That’s where Remark comes in.

The two-year-old startup is helping shoppers buy with more confidence, said Theo Satloff, co-founder and CEO of Remark.

It does this by pairing shoppers with high-quality product experts via an asynchronous live chat with one of 50,000 human experts — artists, musicians, stylists, golfers, ski instructors — who can discuss items just like a retail staffer would do.

And there’s a twist: Remark also trains artificial intelligence models on those experts to create personas that can answer questions with the same style of their human counterparts. This way, the “expert” is always available even if they can’t physically be. Remark also gives the human experts a cut of each sale made through the platform.

Having both a human and AI persona enables the 45 brands already using Remark to tap into the experts, but also into blog posts and landing pages that are optimized for individual shoppers. The company takes a small commission on revenue that's attributed to Remark. There are no platform fees, and the company shares the profits.

“Our objective is to be the best possible guide for them directly in any e-commerce experience,” Satloff said.

Satloff and co-founders Ian Patterson and Carl-Philip Majgaard initially thought expensive purchases, like those skis, would make up most of the use cases. However, they realized that shopping, regardless of price, is an emotional thing, and so customers were using Remark for almost any purchase, like socks.

For Darn Tough Socks, Remark experts spoke with tens of thousands of customers and had hundreds of thousands of interactions with them, all to buy socks.

“For us that was an eye-opening moment because it told us that even something that's pretty low in cost is still a high-consideration, high-emotion purchase,” Satloff said.

Example of Remark's asynchronous chat technology working with Embr's wearables site.
Example of Remark's asynchronous chat technology working with Embr's wearables site.

Remark currently works with customers in the outdoor industry, baby products, beauty and skincare. Customers are seeing a 9% revenue lift and a 30% conversion rate, which Satloff called “astounding.”

He compares it to a brick-and-mortar experience -- when you walk into a store, the retailer has about a 30% likelihood of converting that person to a purchaser.

“Brands are viewing their websites as their new flagship experience, so they need to have that same sort of handheld exceptional guide that they do in brick-and mortar-retail, on their own sites,” he said.

Remark's expert-assisted shopping is a step further than the AI-based algorithmic guides based on past purchases, which were pioneered by companies like Amazon and Intercom with AI-first customer service, and startups like Shoptrue for fashion and Halla for groceries.

Satloff said some of these are more focused on post-sale customer service, while Remark focuses on presale decision support and guidance. The company also has an advantage in that it owns the community, meaning Remark is able to generate new data in perpetuity and on-the-fly as needed.

“We're leading the charge on this persona-based model building, which means that we are literally breaking ground on new techniques and new approaches to role playing and using multiple models in one go,” Satloff said.

To continue developing the product and technology, Remark recently raised $10.3 million in seed funding from an investor group including Spero Ventures, Stripe, Shine Capital, Neo, Sugar Capital, Visible Ventures, and angel investors like Dave Habiger, CEO of JD Power and chairman of Reddit; Jeff Barnett, former CEO of Demandware/CommerceCloud); and Varsha Rao, former CEO of Nurx.

We're launching an AI newsletter! Sign up here to start receiving it in your inboxes on June 5.