Recommendations

Recommendations are a powerful tool in marketing automation, enabling businesses to dynamically suggest products or content to users based on their behavior, preferences, and contextual data. This personalized approach helps drive engagement, improve conversions, and enhance the user experience.

Typical Use Cases for Recommendations

Recommendations can be applied across various scenarios to enhance user engagement and drive results:

• Personalized Shopping Experiences: Display products related to those a user has viewed or purchased.

• Upselling and Cross-Selling: Suggest complementary or higher-value items based on the contents of a user’s cart.

• Category and Brand Promotions: Highlight trending or top-rated products in specific categories or brands.

• Event-Specific Suggestions: Offer seasonal or event-based product recommendations, such as holiday gifts or back-to-school supplies.

Understanding the Key Components of Recommendations

Recommendations rely on several key components: Catalog, Assortments, Blacklists, and Rules. These components work together to define what is recommended, how recommendations are filtered, and the logic used for suggestions.

Catalog

The Catalog is the foundation of the recommendation system. It contains all product-related data that Mapp Intelligence uses to generate recommendations. Product data must be imported and kept up to date to ensure accurate and relevant recommendations.

Key points about the Catalog:

• Product Information: Includes essential details such as product availability, categories, and attributes like size, price, or sale status.

• Data Sources: Information can be transferred via:

  • A data feed (automated import via Mapp Q3).

  • JSON or Excel uploads.

  • Direct transfer from the website.

• Setup in Mapp Q3: Product categories are created or updated in Mapp Q3 > Configuration > Categories > Product Categories. These categories are then imported into Marketing Automation.

Example: If you want to recommend only available shoes in a user’s size, ensure size and availability data are included in the Catalog.

Assortments

Assortments define the pool of products eligible for recommendations.

They allow you to curate and filter products for specific campaigns or contexts:

• Category-Specific Assortments: Restrict recommendations to specific categories.

  Example: For a fitness campaign, show only products in “Sportswear” and “Equipment.”
  

• Custom Attribute Filtering: Include or exclude products based on custom attributes, such as price range or stock availability.

  Example: Exclude products under $20 to focus on premium items.

Configuration Path: Marketing Automation > Contents > Recommendations > Assortments

Blacklists

Blacklists exclude specific products or categories from recommendations, ensuring the user experience remains relevant.

Predefined are:

• Recently bought Products: Avoid recommending items that users have already bought within the last 90 days.

  Example: If a customer purchased a smartphone, exclude the same model from future recommendations.
  

• Recently bought Product - Main Categories: Avoid recommending product caterories that users have already bought within the last 90 days.

  Example: If a customer purchased a smartphone, exclude all smartphones from future recommendations.

Rules

Rules determine how and why products are recommended. They are predefined fall into three main types:

1. Personalized Rules: Based on user-specific behavior.

   Example: Recommend items the user recently viewed, added to their cart, or purchased.
   

2. Association Rules: Suggest items that are often bought or viewed together.

   Example: A customer buying a gaming console might be recommended controllers or games.
   

3. General Rules: Highlight popular or trending products across all users.

   Example: Display “Top Sellers” or “Hot Items” for users with no prior interaction history.

See the detailed explanation in the Recommendation Rules section below.

Steps to Prepare for Recommendations

The following steps are foundational to setting up recommendations and must be completed before creating a campaign:

Step 1: Define Your Catalog

1. In Marketing Automation > Contents > Recommendations > Catalog, select the relevant product categories imported from Mapp Q3.

2. Assign the appropriate datatype (e.g., text, figure, currency, URL) to each category.

3. Ensure product data is up to date by scheduling regular imports via feeds.

Step 2: Set Up Assortments

Navigate to Marketing Automation > Contents > Recommendations > Assortments:

• Create Assortments to define the pool of products eligible for recommendations.

• Filter products based on categories, attributes, or other criteria relevant to your campaign.

Step 3: Configure Placements

Recommendations require specific placements to display correctly. The two supported placement types are:

• OnSite Widget: Configure this placement type to define the layout, text length, and number of recommendations visible. You can also preview the placement, activate automatic scrolling, apply CSS styling, select categories, set up a control group for testing, and manage blacklists.

• Reco API: Enables custom implementations by fetching recommendations programmatically for advanced use cases.

Ensure the desired placement is configured in advance under Marketing Automation > Configuration > Placements.

Steps to Create a Recommendation Campaign

Once the foundational setup is complete, follow these steps to create and launch a recommendation campaign:

Step 1: Create the Campaign

1. Navigate to Marketing Automation > Campaigns and select Create a New Campaign.

2. Define the following elements:

   • Target Group: Choose the audience segment you want to target. More information here.

   • Content: Select the recommendation content you prepared in the foundational setup.

   • Frequency Cap: (Optional) Set a limit on how often recommendations are displayed to the same user.

   • Placement: Choose where the recommendation will appear (e.g., product pages, checkout, homepage). More information here.

3. Save and activate your campaign.

Step 2: Track Performance in Mapp Intelligence

Performance metrics for Recommendations campaigns are available in the Campaigns > Performance tab. Key metrics include:

• Clicks: Tracks how often users interact with recommendations.

• Views: Measures how often recommendations are displayed.

• Conversions: Tracks the number of purchases resulting from recommendation clicks.

• Engagement: Evaluates subsequent page impressions after users interact with the recommendations.

For advanced analyses, click Explore in Intelligence to access:

• Revenue attributed to recommendations.

• Filters to segment performance by category, campaign, or audience.

Available Recommendation Rules

Rules determine the logic behind product recommendations. Each campaign can use only one rule, but multiple campaigns with different rules can be assigned to the same placement. This allows for flexible prioritization and combination of rules to optimize the recommendations shown to users.

Rule Categories and Priorities

We offer a wide variety of rules that can be grouped into three main categories: Personalized Rules, Association Rules, and General Rules. Each category serves a different purpose and has a different value when creating effective and relevant recommendations.

While we recommend prioritizing Personalized Rules highest, followed by Association Rules, and finally General Rules, it is up to customers to define these priorities when setting up their placements and campaigns. Assigning appropriate priorities ensures that recommendations align with user behavior and maximize relevance.

Prioritization Recommendation

1. Personalized Rules (Highest Priority)

   Personalized rules are the most valuable because they are based on individual user behavior and interaction history. We recommend assigning these rules the highest priority in placements.

   • Use Case: Recommending products the user has already interacted with (e.g., added to cart, viewed).

   • Example: A user viewed or added a specific smartphone to their cart. Recommendations would include this exact smartphone as the most relevant suggestion.
     

2. Association Rules (Medium Priority)

   Association rules suggest products based on patterns of user behavior, such as frequently bought or viewed items. These rules are valuable for users with some product interaction.

   • Use Case: Recommending products often bought or viewed together by other users.

   • Example: A user viewed a smartphone. Recommendations might include cases or similar smartphones frequently purchased together.
     

3. General Rules (Lowest Priority)

   General rules apply to users without specific behavior data. While less targeted, they provide a baseline for new users or those without recent product interactions.

   • Use Case: Displaying top-selling or trending products to users without browsing or purchase history.

   • Example: A new user visits the website. Recommendations show the most popular smartphones or trending products.

Assigning Rules to Campaigns and Placements

• One Rule per Campaign: Each recommendation campaign is assigned one rule.

  Example: A campaign using the “Last Seen Products” rule targets returning users based on their browsing history.

• Multiple Campaigns per Placement: To use multiple rules, assign separate campaigns to the same placement.
  Example: A placement includes:

  • A campaign with the “Added to basket” rule.

  • A campaign with the “Last Seen Products” rule.

  • A campaign with the “Top Sellers” rule.

• Priority Handling by Placement: Placements evaluate campaigns based on their priority. Slots are filled by the highest-priority campaign first. If slots remain, the next campaign in priority is evaluated.

Detailed Rules by Category

Personalized Rules

Association Rules

General Rules

Contact your consultant for customizing rules or setting up individual logic.