Product recommendations. 6 best strategies for e-commerce

Customer engagement and loyalty is the goal of every business, its competitive advantage. Product recommendations facilitate the search and shape the customer experience. It is one of the most popular methods of product promotion.

On-page recommendations increase average order value and impact key metrics such as reach and average customer retention.

49% of shoppers order after reading personalised recommendations, even if they weren't planning to buy anything.
52% of consumers are willing to share personal information in exchange for recommendations from a retailer.
54% of retailers believe that recommendations have a direct impact on the average order value.
35% of sales on Amazon are products promoted on the platform based on recommendations.

Recommendation Systems
Referral systems are widely used by e-commerce platforms such as Shopify, WooCommerce, Magento. Built-in CRM systems and Google Analytics allow you to track and analyse user actions, behaviour and preferences. Machine learning algorithms perform collaborative and content-based filtering of purchase history and products viewed, suggesting products relevant to queries and searches. Segmentation based on characteristics, behaviours and interests enables personalised and grouped offers.

1. Bestsellers.
Bestsellers - this strategy recommends popular products in the seller's shop that are in high demand. It is interesting and profitable to create sets of bestsellers of different categories for different types of products. Such recommendations influence buying behaviour, increase unmotivated actions, show benefits and increase sales.

2. New arrivals
New" - this strategy recommends products to customers that have recently gone on sale, been added to a shop or catalogue. It allows you to organically increase the audience reach. In most cases, the new products block is placed on the main page of the site, or the user is directed to a specially created page. For this strategy, the products are sorted according to their date of addition to the catalogue.

3. Related products
This strategy is most commonly used in cross-selling. Once a customer has added a product to their basket, they are offered the same or similar products. By increasing the value of the basket, the merchant increases profits.

It is most effective to place recommendations for related products on the product page and in the shopping basket.

4. Similar items
With this strategy, products with similar characteristics, features and parameters are shown to customers who are placing an order. The strategy can be implemented by placing a 'similar items' section on the website or online shop pages, or by adding a 'related items' or 'often bought together' field to the shopping basket.

5. Recently viewed product.
The recommendation system keeps track of the product pages a user has visited. Personalised recommendations are most often placed at the bottom of the product page, in categories and on home pages. This strategy is based on the assumption that a user's system behaviour can shape their future interests.

6. Based on previous history.
Customers who bought this product also bought...'.
This is a strategy of personalised recommendations based on previous purchases, views and interactions with the store. It is more commonly used in mailings to suggest a product that may complement previous purchases. Such recommendations should take into account changes in customer behaviour and product range, updated in real time.

Halla Systems: from simple to intelligent recommendationsRecommendations can be made both online, for example when adding products to the shopping basket, and directly by calling the customer to clarify the details of the order. Experienced dropshippers and those just starting out are constantly looking for ways to increase both the average check and the total value of an order. Average order (basket) value is one of the most important e-commerce metrics. It is a metric that helps track the average order value of an individual customer in a given time period.

Halla Systems uses both simple (e.g. lists of best-selling products) and complex intelligent recommendations (custom algorithms of dynamically updated product sets) in its work with partners. In the case of recommendations, Halla Systems helps dropshippers select products relevant to their queries by applying different search algorithms, absolute and relational filtering.