The average shopping cart abandonment rate is a staggering 70%. As a result, preventing lost sales is a main priority for ecommerce retailers. What's more, the apparel industry is highly dynamic due to item seasonality, customers' unique fashion tastes, and the ever-evolving definition of what's in style.
It's important for retailers selling online to offer the most relevant item results – and the increase in mobile shopping only reinforces this need. Mobile commerce sales grew 41.4% in 2020, and is expected to grow another 15.2% this year, according to eMarketer.
Whether on a browser or in a retailer's app, the limited number of items displayed on a smartphone demands retailers provide customers with relevant, personalized results to reduce the number of lost sales.
Image similarity, or product recommendations calculated on the basis of patterns recognized between images, can improve and influence the shopping experience on a product webpage, in the shopping cart and on a shopper's wish list.
1. Image similarity on the webpage
After clicking on a product that strikes their interest, a customer is taken to the item's unique webpage. Here they can find more information about a product and gather a detailed view, including various images and videos.
Many retailers place additional product recommendations under the item details based on different recommendation strategies. These recommendations are calculated on the basis of similarity analysis.
For example, a customer may receive suggestions for a shirt that is similar to the one they originally selected in terms of fabric, material and/or color. In the event their desired item is sold out in size, instead of a shopper leaving the site, retailers can share similar items that might still fit a customer's needs.
2. Image similarity in the shopping cart
Once a customer has decided to buy an item of clothing, they place it in their online shopping cart. From here, they should ideally not be distracted by anything, to ensure they do not abandon the purchase. However, we know this is not a flawless strategy. A customer may open a new tab, or physically step away from their device. If items are selling quickly, a customer may return to their shopping cart only to find the item is sold out.
To counteract the threat of lost sales, it has proven practical to display alternative products that are visually similar to the sold-out item of clothing. Retailers can then define the filter criteria as it best fits the customer's need. For example, they can ensure that the items displayed are available in the clothing size of the sold-out product.
3. Image similarity on shopper wish lists
A wish list makes it easy for customers to remember items they may want to buy later. Apparel retailers often see a wish list as a preliminary stage to the actual buying process. Unlike in the shopping cart, where the customer should not be distracted, opportunities for upsell exist in the form of recommendations.
For example, recommendations can be displayed based on visually similar items of clothing that are on a wish list. Using artificial intelligence, retailers can analyze which recommendations are more frequently accepted by customers and added to wish lists – or shopping carts, too – to better inform future merchandising decisions, such as the choice to display items of clothing within a certain price range based on the price of the original product.
Ecommerce continues to rise
As the popularity of ecommerce continues to rise, retailers can improve the shopper experience from the first click all the way to the point of purchase. With image similarity, retailers can achieve higher sales by offering more accurate product recommendations.
prudsys, a member of the GK Software Group, works to automate personalization in retail through artificial intelligence. Learn more about prudsys solutions here.
