Michelle Bacharach says she spent years feeling style-challenged: She’d cheerfully purchase an item that caught her eye, only to feel frustrated later when she had no idea how to pair it with other items in her wardrobe.
While most of us hang items like that in our closets and endure pangs of guilt every time we notice them, Bacharach developed a more ambitious plan. She created FindMine, a system that uses machine learning and deep data mining to take the friction out of “completing the look.” It chooses products from a retailer’s inventory to make recommendations that help consumers create an outfit, build out a room’s décor or understand how to put together a skin care regimen.
“When you show customers how to use what they’re buying, they will spend 200 percent more,” says Bacharach, co-founder and CEO of FindMine. “Retailers and brands have known this forever — when they merchandise outfits on mannequins, those pieces sell out quickly. They know that shoppers are looking to them for expertise and guidance and will spend more when they receive it. The problem is, not every associate is a stylist and not every retailer has the deep pockets to support a large staff.”
FindMine allows retailers to share their brand’s point of view about using products, Bacharach says, “then apply it across the entire product catalog and across every single touchpoint in the customer’s potential journey.” She says by showing full outfits and suggesting items that “complete the look,” FindMine can increase average order value by up to 150 percent. In addition, the amount of time shoppers spend exploring products doubles when they can see how to put an outfit together.
FindMine’s algorithm creates over 11 million outfits daily across ecommerce and mobile websites, marketing channels like Facebook Messenger, email campaigns and stores. The technology uses artificial intelligence to create outfits from a brand’s full product catalog in the same way a personal shopper would, but it does it at scale. For shoppers, it’s like having a personal stylist. For retailers, it’s an opportunity to increase basket size.
A dozen fashion brands are using the technology, including Perry Ellis, John Vavartos and Laundry by Shelli Segal. On average, FindMine users experience a 6 percent incremental revenue increase.
“We’re using technology to help scale out the content creation that a retailer would otherwise have to do by hand,” Bacharach says. “Over time, the technology ‘learns’ when it’s right and when it’s wrong, and the constant feedback reinforces our algorithms making them better and smarter.”
Before a retailer can flip the switch on FindMine, every product is added to the system and the models and algorithms which have been created are applied. Merchants typically have full access to the content management system, so they can see the outfits being created and make edits. “FindMine uses data to determine the best outfit to suggest to a customer — and it shows them items that are in-stock and seasonally appropriate,” she says. “It can be a huge time saver for merchants.”
It can also deliver a lift on sales. Bacharach says retailers can blend FindMine data with customer relationship management information for targeted messaging: Knowing a shopper purchased boots last fall, a retailer can send messaging showing how to wear last year’s purchase with a new outfit for this season. Others are exploring follow-up email campaigns that acknowledge a purchase and provide a 10-day style guide suggesting different ways to wear it.
“In the end, the technology really accentuates the value of customer service,” Bacharach says. “It’s more than just a recommendation — it’s about feeling confident in your choices, feeling valued and nurturing loyalty.”