We are happy to share about the solution

Apr 15, 2020
The evolution of PDP recommendations in fashion
Amazon-style collaborative filtering recommendations on ecommerce websites have been a real breakthrough for online shopping in the past and have helped to increase sales and revenue for many retailers. Amazon has claimed that around 35% of its revenue comes from purchases that have been recommended to customers on product detail pages. But while this strategy might work for Amazon and similar sites, it isn't always effective across other industries. We live in an era where customer experience is more important than ever.

In order to match customer expectations, retailers need to personalize every step of the purchasing journey.

Recommendations under headings such as 'you may also like' can work in some industries, for example electronics, when there are clearly defined parameters that suggest other relevant items for the customer to purchase. Fashion is another matter entirely. Instagram and influencers are now promoting the latest trends and everything moves so quickly. Style is so subjective and a matter of personal taste – if a customer sees the same products displayed on every page and doesn't like that particular style, this method of displaying other products is practically useless. It will not encourage the customer to browse the catalogue further, or to purchase additional items that the customer really likes. This method makes the customers feel like the retailer doesn't care about them, doesn't understand them and makes the online shopping experience impersonal. In the highly competitive world of online fashion, this can be a mistake: if your customers go elsewhere, they might not come back. 
There is, however, another way entirely to offer product suggestions to your customers. That is to make additional purchase suggestions that carefully suit their needs and offer them perfectly matching items. Instead of just selling individual pieces, you give your customer an experience that makes them feel special – an additional styling service for free. 'Entire outfit suggestions' both personalize the service for the customer and let the retailer provide more upsells using a smarter tool. 


 
Companies like Inditex, Farfetch and Asos are already generating up to 5% of their annual revenue by using 'complete the outfit' suggestions on their product detail pages. 

This is a staple for many fashion monobrands. They often have their own unique brand styling strategy and employ numerous designers, stylists, merchandisers, photographers, models and post production team to consistently and seamlessly showcase the brand across all platforms. They put together completed outfits that are shown in their catalogues, websites and social media and don't need to create any other recommendations. 
 
However the economic realities for multi-brand marketplaces is entirely different and they have leaner budgets and don't have the luxury of employing these teams of people to style and promote their products. Marketplaces use a combination of their own photos as well as the photos provided by the labels and brands that they stock. When this happens, there is often no seamless way to offer matching or complementing items.
As customers are unique and come from a range of different demographics with different style preferences, recommending only one outfit means you only appeal to a narrow percentage of the audience.  On the other side of the coin, creating and photographing many outfits manually to make recommendations will be costly and not a wise business decision. The solution to provide a variety of outfit recommendations tailored to customers' unique preferences, with a positive ROI, is to use artificial intelligence. 
 
SuitApp is an AI software solution for fashion retailers that can be embedded into websites to provide customers with limitless suggestions about outfit choices, increasing the average sale size. Recommendations are made by a range of factors such as style, color, fabric, textures and other parameters. It can also predict trends and help to manage inventory to be more responsive to customer needs.