Personalized Product Recommendations in Liferay Commerce

For anyone launching an e-commerce site today, personalized product recommendations are an obvious feature to include. However, these automated experiences are not as simple as e-commerce leaders make them appear.

Once you bring in real customer data, you have dozens of ways to recommend products to your users. For B2B merchants, there are three types of recommendations that are quick wins for improving the customer experience — and all three are provided out of the box in Liferay Commerce.

Types of Product Recommendations in Liferay Commerce

Liferay Commerce provides machine learning models for implementing personalized product recommendations at three different levels.

Content-Based Product Recommendations 

Help customers find the right product by recommending related products based on product content such as specifications and categorizations. As you sync your product catalog from your ERP or PIM, Liferay Commerce refreshes its recommendation engine to ensure your customers can discover the right related products.

Also-Bought Product Recommendations 

Encourage cross-selling by recommending products that other customers bought, based on the product that a customer is currently viewing. Also-boughts are a popular way to help customers make the right choice when buying a product, as they can easily see what other customers liked and dig further without searching the catalog via keywords or filters.

User-Based Product Recommendations 

Offer more personalized recommendations by showing customers what other customers who are similar to them also bought. This adds known customer profile data to your Also-Bought recommendations for an even more personalized experience, equipping your customers with the popular buying choices from other customers from similar companies.

Where to Start

Based on your strategy, it might be beneficial to start with one type of recommendation that will have more direct impact on your goals. Here are some valuable strategies that B2B merchants can implement to address common pain points.

Increase Visibility of Low-Selling Products

Most catalogs have products that end up in a Miscellaneous category or otherwise low-visibility area of the site. By implementing content-based product recommendations, you can put these products in front of customers more often, increasing the number of eyes on them and hopefully resulting in a sales lift.

Reduce Returns

Many product returns are avoidable, especially if your product lines involve compatibility requirements (e.g., spare auto parts, A/V equipment, software, etc). Pair content-based recommendations with manually-added product relationships to ensure that customers are finding the correct product, ultimately reducing returns.

Speed Up Purchasing Decisions

B2B purchasing cycles can be particularly long as purchasing often involves a higher investment and more complex products. Make the decision easier (and thus faster) by leveraging both also-boughts and user-based recommendations. In addition to ensuring that customers are finding the most relevant product, you can add an additional layer of credibility with the knowledge that other companies have made the same purchase.

Up-Selling for New Accounts

If you know one of your customer segments is likely to purchase your more expensive products, then you can implement user-based product recommendations to ensure that your newest customers within that segment see those products immediately. When you onboard a new customer, implement user-based recommendations instead of regular also-boughts so that they see the big ticket products first. Certain audiences and buyer profiles are more likely to go for premium options if they know their peers are doing the same and you can capitalize on this with the right recommendation during the purchasing process.

There’s More to Learn

These are only a few of the types of product recommendations you can create with your customer data. Once you bring in specific customer data such as location, company size, device, buyer role and so on, you can create incredibly personalized recommendations to address many needs within your strategy. The key is to think beyond marketing personalization — banners, coupons and ads — to personalization that delivers higher-value offerings to your customers, aligning with the way they want to buy products.