Christmas holidays vibes are in the air! And online retailers try to make the most out of the season by offering promotions and finding their perfect mix.
But they are always faced with two major problems: site abandonment and especially high return rates.
Customers can leave an ecommerce website for many reasons. But we shouldn’t disregard how shocking the numbers are: typical ecommerce abandonment rates range from 50% to 80%. The retail industry, in particular, suffers from an average abandonment rate of 72.8%. Clothing category its hit by a 65.9% rate.
Consumers leaving shopping sites without making a purchase is a huge burden for brands and retailers. In fact, it costs them between 2 and 4 trillion per year.
Pre-product page abandonment
We could say there are two types of eCommerce visitors. There are those who use the site search functionality, and those navigate through the web using menus and banners. The portion of shoppers using the search function is way smaller. In fact, only around 20% of visitors make use of site search. However, this exact portion is usually the one with a higher purchase intent.
These visitors are 3 to 4 times more likely to convert on your website than shoppers who don’t use search. If they don’t make it to the product page, then site search is not as good as it should be.
Top potential problems to look at would be:
- Wrong product classification in different categories
- Lack or insufficient product images on category pages
- Poor filtering functionality
- Unclear layout that leads to confusing merchandising
Reducing abandonment and driving conversions
Providing consistent and comprehensive product information is key to enable top-notch product filtering and highly relevant site search results. These two functionalities act as important traffic drivers to product pages. But we all know this can be costly and time-consuming.
Automated product tagging uses Artificial Intelligence to classify and assign relevant and fine-grained attributes to fashion products based on their images. This improves product SEO and makes items more easily discoverable within your website.
We brought them to the product detail page. Now what?
Once in the product page, the shopper benefits from more complete information. This drives higher conversions, as 98% of consumers feel discouraged to make a purchase when product details are incorrect or incomplete.
The shopper still might not want to buy that specific product due to many reasons, for instance:
- The customer feels the item is too pricey.
- The size the customer is looking for is no longer in stock.
- Product images don’t have enough quality or don’t show enough details.
- The style of the item does not exactly match the customer’s taste.
To tackle some of this issues, we turn to AI once again. A solution based on machine learning and computer vision can automatically suggest items that look similar to the one the customer is looking at. Those suggestions will show items with similar features but being slightly different. In turn, this will increase the chances that customers find exactly what they like in their exact size.
Fighting eCommerce return rates
eCommerce return rates are traditionally much higher than return rates in brick-and-mortar stores. In fact, 89% of buyers have returned an online purchase. Obviously, this happens largely due to the fact that customers can’t touch or try products when they buy online. Actually:
- 41% of customers buy the items they like in multiple sizes and return the ones that don’t fit them.
- 77% of returns come from recurring customers
In the US, returns in online sales have been reported to account for 20% to 40% of sales. The number one reason for returns is usually a poor or incorrect fit. This is why some retailers like Asos have already implemented strategies like virtual fitting solutions.
But there are other important reasons for returns:
- Up to 40% of customers say they have returned a product due to a mistake or lack of product information.
- 22% of returns are due to the fact that the product looks different in real life.
The first factor can be tackled by generating proper product information through AI, as we have previously mentioned. The second one, is usually due to bad quality images. This is a real pain for retailers, and even more for marketplaces, that receive images from different providers lacking consistency.
However, deep learning techniques have managed to give a helping hand to online retailers. How? They automatically detect images that don’t have enough quality or don’t show all necessary product views. This is a good way to realistically show customers how the product looks like. And by providing the highest definition and accuracy, return rates are reduced.