The renowned furniture retail group Schaffrath recently started using our Visual Search AI for product recommendations and visual product searches. The case shows how online shops can make more out of a neglected resource: their product images.
The company:
Schaffrath is one of the leading furniture stores in the Rhineland. The company operates 16 furniture stores and kitchen studios on the Lower Rhine and Cologne. In its online shop Schaffrath.com, the retailer offers around 14,000 products: furnishings, garden products, decoration, accessories, and household items. Schaffrath wants to expand this range in the coming years – and has therefore decided to use our technology.
The challenge:
Every online retailer knows this problem: the new customer campaigns work, but the expensive traffic only leads to relatively low conversion rates.
Schaffrath faces this challenge to convert a maximum rate of online visitors to customers. The basic idea: customers should need as few clicks as possible to find the relevant furniture or home accessory on Schaffrath.com. Because: "Every additional click prolongs the buying process for the customer and potentially costs us sales," says Schaffrath e-commerce division manager Thomas Langen.
The solution:
Since the summer of 2023, our Visual Relevance Machine Vviinn has been integrated into Schaffrath's webshop. Our AI now ensures that product recommendations are more precise and better – and also helps to reduce unnecessary search clicks.
Imagine the following scenario: A customer is looking for a new vase for her dining room. She quickly finds an appealing model on the list. But she wonders what else is available in this style. She clicks on the recently introduced search box below the product image, and a selection of stylistically similar models immediately appears. Now the vase of her dreams is there: an elegant grey model that fits perfectly with the glass cabinet in the dining room.
Our image analysis method based on artificial intelligence (AI) delivers significantly better results than conventional text-based search systems. Instead of having to click through many text filter options, our Visual Relevance Machine captures the image the customer has in mind much more quickly and accurately when searching. We call this true personalization.
Our fictional customer could also use our new image search on Schaffrath.com. A vase caught her eye on the Instagram page of a home furnishing magazine. Using the camera icon next to the search box, she uploads the picture directly to Schaffrath.com. A selection of stylistically similar vases appears on the display. And indeed: Schaffrath has a very similar model in its range.
The customer could use any picture. Self-shot photos from the smartphone or even drawings - our AI recognizes and processes everything.
In addition, Schaffrath also uses our Vviinn technology for cross-selling. In addition to "similar products," users receive "matching products" recommendations. For example, we show the person interested in vases and other household items, such as a matching coffee service.
Head of e-commerce Langen is delighted with the first results. "After a few tests, we saw the results were excellent. Especially the great speed of Vviinn is a big plus."
We are pleased to have found a partner in Schaffrath who is so open to innovation and allows us to show our technology's performance. We look forward to further cooperation.