Recommend Perfect Products with MoEngage’s Smart Recommendations

Reading Time: 9 minutes

Did you know that 35% of Amazon’s E-commerce revenue is generated thanks to its advanced product recommendation engine? Did you also know that 80% of the stream time on Netflix can be attributed to its industry-grade content recommendation engine?

The modern consumer wants more personalized and contextually relevant communication. 27% of consumers find irrelevant content and product recommendations to be their most frustrating experience with brands, while 26% of consumers want brands to personalize experiences based on previous shopping history.

Customers no longer question the need to collect data because they expect you to use it to create delightful experiences for them. This includes delivering personalized recommendations to them. Although consumer brands understand the need for personalized recommendations, most are not well-equipped to create memorable customer experiences.

We’ve built Smart Recommendations to help you build personalized micro-moments for your customers. In real-time, you can send suitable recommendations to each customer based on their preferences, past interactions, and engagement patterns, easing product discovery and increasing conversion.

AI powers the all-new MoEngage Smart Recommendations to enable you to send real-time recommendations across any channel without any technical expertise.

This helps your customers discover new products and increase their probability of completing a purchase. You can create a seamless buying journey personalized to each individual’s unique needs by recommending the right products and items.

What Are the Different Recommendation Models You Can Use?

Smart Recommendations has six distinct built-in recommendation models, enabling you to serve your customers the most appropriate product recommendations.

1. Recommended Items

Leverage our AI-powered algorithms to recommend the most suitable products for your customers.

This model filters items based on customer preferences, past interactions, and other customers’ engagement patterns, particularly from the last two months, to suggest products that are most relevant to them.

For example, if a customer checked out fitness shoes lately and prefers a certain brand, the model might recommend a pair of sports shoes accordingly.

If the customer is new or has been inactive, the model uses a fallback mechanism, suggesting trending and popular items that align with the customer’s captured characteristics—such as preferences, demographics, and other properties.

For example, a fitness enthusiast male from Canada might receive a recommendation for a trending men’s winter workout jacket, as shown in the image below.

This recommendation model is ideal for maximizing engagement, driving more purchases, and delivering personalized experiences.

2. Trending Items recommendations

You can use this recommendation model to boost conversions by recommending products that have garnered social validation and to guide new customers to a purchase on your website or mobile app.

The Trending Items recommendations model uses MoEngage’s proprietary AI, Sherpa, to recommend the most popular items products based on historical customer actions such as the frequency when an item was viewed, wishlisted, purchased, or added to cart.

A few examples of Trending Items include recommending the most viewed products in a particular month (like Sony’s Playstation 5 in the month of its release) or the best sellers in a specific category (like the Apple iPhone 15).

3. Item attributes-based recommendations

You can use this model to recommend products based on specific item (product) attributes.

This model is best used when you want to showcase products for specific thematic events, occasions, requirements, and more.

Some examples of item attributes-based recommendations are recommending skirts that are “blue” in color and “large” in size or suggesting mutual funds that have “low SIP installments,” “high-interest rates,” and a “1 yr maturity period.”

4. Customer action-based recommendations

You can recommend products based on your customer’s past interactions and actions.

Customer action-based recommendations play a critical role in driving purchases for items your customers want.

Some examples include recommending products added to the cart but not purchased or suggesting TV shows customers have added to the wish list.

5. Similar items

Let Sherpa AI recommend products similar to those a customer has last interacted with.

This model picks items based on product attributes and the historical co-occurrence of events.

The similar items recommendation model will help you avoid and reduce drop-offs in the buyer journey.

An example includes recommending dresses based on attributes such as price, brand, title, or category, often viewed together, added to a cart together, and more to a customer who has viewed a winter dress.

6. Frequently viewed together

Leverage Sherpa AI to recommend products that customers frequently view together. This will depend on the latest product viewed by a customer.

This model is critical when you want to avoid and reduce drop-offs and showcase catalog versatility.

An example of this model being used is when a customer has viewed a phone, you can recommend other products, such as phone cases or accessories frequently viewed together.

7. Frequently bought together

Use Sherpa AI to recommend products frequently bought with the latest producing listing a customer has interacted with.

This model is a great way to boost cross-sell opportunities.

An example of this model is recommending accessories other customers have frequently purchased to a customer who purchases a Television unit.

Why Did We Build Smart Recommendations?

Personalized recommendations have become necessary for brands to engage with customers, boost sales, and deepen engagement. 49% of customers had purchased a product because of a personalized recommendation they received from the brand.

We built Smart Recommendations to help brands send personalized, contextually relevant recommendations to customers.

This new feature enables you to send real-time, contextually relevant suggestions from millions of catalog products to each customer. The new Smart Recommendations will help you:

1. Ease product discovery
   MoEngage Smart Recommendations empowers brands to cut through the clutter and deliver the right products to customers when they’re looking for them.
2. Prevent customer dissatisfaction
   It can be a little annoying when customers search for products but cannot find them. When this happens regularly, it creates unhappy customers, possibly a bad reputation for the brand, and might lead to churn. Use Smart Recommendations to ensure that customers don’t have unhappy product search experiences.
3. Create the perfect purchase journey
   Each customer follows a different path to the purchase. Leverage Smart Recommendations to personalize the journey for each customer across multiple channels.
4. Deliver better shopping experiences
   49% of customers say they purchased a product they didn’t initially intend to buy after receiving a personalized recommendation. Create tailored experiences for your customers using Smart Recommendations and increase trust in your brand.

What Sets Smart Recommendations Apart From Other Solutions?

Smart Recommendations make it easier for your customers to discover new and relevant products. Here are reasons why our MoEngage is the solution to your current challenges:

1. Built-in recommendation models for every use case

Your customers’ needs are varied, and so are the use cases for your brand.

Marketers can choose the right recommendation model to implement the most appropriate strategy and delight customers with relevant recommendations – Item attributes-based, Customer actions-based, or AI-recommended.

2. AI-powered recommendations based on historical data

Leverage the power of AI to recommend products that would be most relevant for the customer.

Our recommendation engine keeps track of all customer interactions, preferences, and engagement patterns in real time. This information is then used to decode your customer’s purchase intent and deliver the most relevant recommendations.

3. Refine recommendations using filters for maximum relevance

With Recommendation Filters, you can now refine suggestions based on customer actions and product attributes to ensure you recommend only relevant products.

• User Action-Based Filters: Refine recommendations based on customer actions. For example, suggesting similar products and excluding already purchased items.
• Item Attribute-Based Filters: Refine recommendations based on product attributes. For example, suggest frequently bought together items that are either on sale, or from a specific brand, or available in the customer’s city.

These filters help you suggest precise and relevant recommendations for your customers.

4. Real-time recommendations

MoEngage’s recommendation engine feeds customer behavior to our proprietary algorithm as it happens, adapts to the new data, and refreshes quickly to provide relevant recommendations in real time.

5. Access to the recommendation technology that powers Amazon

MoEngage Smart Recommendations is built using AWS Personalize, the same recommendation engine that powers Amazon’s robust e-commerce website.

You can be assured that you get the most accurate and relevant recommendations powered by industry-leading and battle-tested algorithms.

6. Leverage all available communication channels

Delight your customers with relevant product recommendations across the website and the channels your customers prefer, such as Email, Push Notifications, SMS, In-App Messages, On-Site Messages, and Cards.

7. Go live without any technical expertise

Marketers don’t need data science or coding expertise to go live with Smart Recommendations.

You can integrate data, select recommendation models, and start serving relevant product recommendations quickly, all from an intuitive and easy-to-use interface.

How Can Smart Recommendations Benefit Your Brand?

MoEngage Smart Recommendations lets you send personalized recommendations to your customers across different touchpoints in your customer’s omnichannel journey, such as Push Notifications, Emails, WhatsApp, Facebook, Mobile In-app Messages, and Website Banners. You can personalize the recommendations for each customer journey stage and enable your customers to find what they want quickly. Your customers no longer need to click or search multiple times to find their favorite product on your website. This will help you:

Boost Revenue

Suggest the most relevant recommendations to each customer based on their preferences, helping them discover products and services faster. By helping your customers fulfill their needs more quickly, you can improve the overall experience and increase revenue.

Increase Average Order Value and LTV

Recommend popular, relevant, and trending products to your customers when they’re about to complete a purchase and encourage them to increase their order value. Over time, this also helps you improve customer LTV.

Improve Conversion Rates

Customers are more likely to purchase products they resonate with. So if your campaigns recommend products aligned with their preferences, they are more likely to convert and give you better conversion rates overall.

Reduce Cart Abandonment

Suppose your customers are hovering around the exit button or performing other actions that indicate they’re uninterested in completing a purchase. In that case, you can slip in a relevant recommendation to pique their interest. This helps reduce cart abandonment and improve conversions.

Create Better Experiences

With MoEngage Smart Recommendations, you can show your customers exactly what they’re looking for before they even start searching. Surprise your customers at every step with a relevant, personalized recommendation and create better experiences for every customer.

Influence Usage and Drive Retention

Recommend the newest shows or songs from a customer’s favorite artist. Recommend precisely what customers want before they search for it and improve their trust in your brand. These recommendations will also help you create happy customer experiences that keep them coming back for more.

Everyday Use Cases for Smart Recommendations

Retail and E-commerce

• Brands can leverage the AI recommendation model to suggest clothing items best suited to customers based on their preferences and past interactions.
• Brands can also remind customers to complete the purchase of products they’ve abandoned in their carts.
• Brands can recommend specific products to customers during special occasions, such as “red” colored “hats” with “cotton” fabric during Christmas.
• If a customer views a pair of running shoes of a certain brand, you can recommend similar running shoes of the same or a different brand with similar pricing.
• When a customer views a TV of a particular brand, marketers can recommend TV units of other competing brands or show accessories such as speakers or streaming devices often viewed together by other customers.
• If a customer purchases a laptop, you can recommend other items, such as a laptop bag or external hard drives often bought together by other customers.

Banking and Financial Institutions

• Brands can leverage AI to recommend credit card point redemption offers best suited to customers based on their preferences and past actions.
• Banks can recommend mutual funds with features specifically relevant to students, such as “low SIP installments,” “high-interest rates,” and “small maturity period.”
• If a customer views a credit card with lifetime free rewards, you can recommend similar reward credit cards with similar benefits.
• When a customer views the available travel insurance of a particular brand, marketers can recommend insurance plans from competing brands or similar products, such as a Forex card, often viewed together by other customers.
• If a customer purchases a life insurance policy, you can recommend other items such as disability insurance, long-term insurance, or retirement plans often bought together by other customers.

Music and OTT Platforms

• Brands can leverage the AI recommendation model to suggest TV shows customers would be most interested in watching based on their preferences and past interactions.
• Brands can recommend “newly arrived” TV series or music albums of a specific genre.
• Brands can suggest “movies” from the “RomCom” genre with a specific actor, such as “Julia Roberts,” during Valentine’s week.
• Brands can also suggest TV shows customers have added to their ‘watchlist.’
• If a customer views a certain movie on your streaming app, you can recommend similar movies with the same actor, director, or genre.
• When a customer explores a specific fashion documentary on your streaming app, you can recommend other fashion documentaries or movies based on real life that other customers often view together.
• Suppose a customer purchases a premium ad-free music subscription. In that case, marketers can recommend other paid add-ons such as offline listening, high-quality audio, or the option to listen together, often bought together by other customers.

How Can You Get Started?

If you’re an existing customer, contact your favorite MoEngage account manager to get started with Smart Recommendations. If you’re new to MoEngage, you can request a demo here.