Collections

Collections lets you build dynamic data driven criteria from your catalog using flexible filters and logic; powering highly contextual, personalized recommendations

As of now you can create your own strategy by manually selecting events. Keep reading to know how it’s done.

Accessing Collections

• Once you’re on the Recommendations page on your dashboard, proceed to click on the 'Add Recommendation' button on the right-hand side of the ‘List of Recommendations’ page.
• Now, on the Template Library Page, scroll down or click on the ‘Collections’ button.

• ‘Create’ card in the 'Collections' section.

How to create Collections?

Begin building your targeted collections strategy by manually building your own filters with our dynamic criteria builder

• Once you’re here, you’re introduced to the create recommendation page, from here you can start building your collection.
• Firstly, start by choosing the Catalog you want to create this collection for from the dropdown.
• Choose a catalog column that you want to sort your criteria by and the sorting order i.e. Ascending or Descending order in which you want your recommendations to be showcased.
• Proceed to build the catalog filter criteria.
• You can rename your recommendation as by default the name is provided based on the date it is created on.

Build Catalog Filter Criteria

Here’s how to create advanced filtering logic using nested AND/ OR conditions. This feature empowers you to precisely target catalog columns for personalized recommendations and campaigns.

Understanding Nested Logic

Nested logic allows you to combine multiple filtering conditions using AND and OR operators, creating complex criteria for your Collections. This is particularly useful for segmenting your catalog based on intricate user behaviors and attributes.

Steps to Build Nested Logic

Left Side (Catalog Column Selection):

• Select the catalog column you wish to filter from the dropdown menu. This column represents the attribute you'll be applying the filter to (e.g., "Product Category," "Price," "User Rating").

Middle Section (Comparison Operators):

• Choose the comparison operator that defines the relationship between the selected catalog column and the comparison value. Available operators include: equals,not equals, greater than, less than, between, contains and other relevant operators depending on the data type of the catalog column.

Right Side (Comparison Values):

• Define the value against which the selected catalog column will be compared. This value can be sourced from:
  • Last Event Attributes: Attributes derived from the user's most recent event (e.g., "Last Viewed Category").
  • User Attributes: Pre-existing user data (e.g., "User City," "Purchase History").
  • Other Catalog Columns: Compare against values within another column of the same catalog (e.g., "Compare Product Price to Retail Price").
  • Constants: Manually entered, static values (e.g., "Red," "100").
  • Campaign-Supplied Dynamic Values: Values passed from campaigns, enabling real-time personalization.
• Adding Multiple Conditions: Click the ➕ button to add additional filtering conditions. This allows you to combine multiple criteria using AND and OR operators.
• Nesting Conditions:
  • WebEngage supports up to 3 additional levels of nested AND/OR logic. This enables you to create highly specific and complex filtering rules.
  • Example: (Category = "Electronics" AND Price < 500) OR (Category = "Clothing" AND (Size = "Large" OR Color = "Blue"))
• Key Considerations:
  • Data Types: Ensure that the comparison values and catalog column data types are compatible.
  • Operator Logic: Carefully consider the use of AND and OR operators to achieve the desired filtering outcome.
  • Nesting Complexity: While 3 levels of nesting are supported, prioritize clarity and maintainability in your filtering logic.
  • By utilizing nested AND/ OR conditions, you can create highly targeted Collections, enhancing personalization and campaign effectiveness.

👍

Use Cases

👉E-commerce & Retail:

1. Personalized Product Suggestions – Show fashion items based on a user's past purchases, preferred brands, and price range, sorted by popularity or discount percentage.
2. Trending Electronics – Filter the latest mobile phones based on brand preference, budget range, and customer ratings, sorted by newest arrivals.
3. Grocery Recommendations – Recommend fresh produce or daily essentials based on past purchases and seasonal availability, sorted by best-sellers.

👉EdTech & Online Learning

1. Course Suggestions – Recommend courses based on a user’s career interests, past enrollments, and available discounts, sorted by completion rates.
2. Book Recommendations – Filter books based on user preferences, reading level, and format (audiobook, ebook, paperback), sorted by highest-rated.

👉Dining and Food Delivery

1. Restaurant Recommendations - Recommending Restaurants in the users city with the users preferred cuisine, within a certain price range, and sorted by average rating.
2. Order Recommendations - Filter meal options from restaurants based on dietary preferences (vegan, keto, gluten-free), sorted by calorie count or rating.

Defining Variables within a Campaign

To define variable you have set in the criteria or as code in the html template, you can follow the instructions below. For example:

{{recommend["c_recommendation_1"][xargs(var1=user["custom"]["custom_city"],var2=user["custom"]["custom_boolean"])]["0"]["city1"]}}

Where "c_recommendation_1" is the Reference ID you can find in the preview or the recommendation listing page.

While creating your campaigns, in the message creation tab you can use your recommendations as variables for the content to use in your campaign as shown below.

Additionally, for complex use cases where you require values from multiple sources to flow into the campaign, you can use:

{% set Day = "now" | date("dddd") %}
{% set Month = "now" | date("MMMM") %}
{% set DayMonth = Month + " " + Day %}
{{ recommend["c_recommendation_1"][xargs(var1=DayMonth)]["0"]["Order"] }}

Here you can setDayMonthas a variable which pulls in two other values, and send it as a variable to the collections query.

For additional support with configuring variable in Campaign for collections, please reach out to your CSM or [email protected].