AI Product Recommendation Engine for E-commerce: How to give online clients an in-store experience

The boom in online shopping 

In the past 10 years, online shopping has become more and more prevalent, but even more so now with the current COVID situation. Online shopping has become the battleground for retailers and the stores offering a good shopping experience will outlast those that don’t. 

Meeting a client’s expectations

Choosing between thousands of products is not easy for clients and for online stores, offering the right products at the right moment is a difficult task. When a client is browsing for style-matching products, they need to manually browse through dozens of brands. Without the right guidance, they can easily get frustrated and leave for another store providing a better shopping experience.

Clients’ expectations have been raised by using platforms such as Netflix, YouTube, or Amazon where they are recommended what to buy. These recommender engines work excellently when a client is regularly using the business, as they build recommendations based on previous data. However, what happens when there is no historic data? This is often the case for clients purchasing new a new bathroom, kitchen, or living room furniture. 

A new type of recommender system

When approaching these kinds of less regular purchases, traditional recommender systems are not up to the job. What people are looking for is items that match in terms of “style.” This is the job of an on-floor salesperson who has a deep knowledge of all the items across all the brands in stock and could welcome clients, find out what they want, and recommend matching items. Clients would leave the shop with a smile on their face avoiding the frustration of searching the entire catalog but rather allowing an expert to advise. 

However, these experts are hard to come by, and with online shopping becoming the norm how do you replicate this experience for people sitting at home and looking for style-matching products, with no historical data, and with a huge stock to choose from? How can you offer the personalized experience they demand?

Using estylo. 

estylo – revolutionizing recommendations 

estylo uses cutting-edge computer vision AI technologies to map all the products in your stock and classify them by distance according to style. This way, a client choosing one product will be offered similar looking products from different vendors so they can make their choice based on style, budget, availability, and other factors. However, estylo doesn’t stop there, it will also offer similar-looking, or style-matching, products from different product categories. This way clients can easily choose different products, which are similar in style to the ones they are browsing, but which they may not have thought of before.

The result is that clients can find what they are looking for very fast, and add extras with a similar style, thus having a great shopping experience. 

Improving business

estylo can improve your business in many ways:  

  • Scale-up: there are no limits to the amount of real-time recommendations estylo can perform. It’s like having an expert digital sales professional, available 24/7. 
  • Increased basket size: clients will be offered style-matching products in other product categories so clients can purchase items they may not have even thought of.
  • Efficient inventory management: by having all your product stock classified into style distance, your products will automatically be offered at the right moment. Clients will then give you interesting insights about trends in demand which you can use for even more efficient re-stocking.
  • Easy to use: estylo is very easy to implement in any store via REST/API with minimum integration resources needed. Therefore, it will save you money and resources in the long run. 

Find out more about how estylo is changing the way people shop by booking a demo with us now

More info at