The recommendation engine for automatic email marketing uses class names, categories, and instructor names in conjunction with each student's registration history to craft customized recommendations for upcoming classes.
Below are some tips to optimize your catalog to ensure these recommendations are as relevant as possible to your students' interests.
Use clear, descriptive class names.
It is best if class names do not include class numbers, season codes, or other identifiers (unless it is absolutely necessary for the program). No need to fill your class names with keywords, though! Keep it simple and readable, and the recommendation engine will take care of the rest!
Use the same class names in each season/catalog.
If you run the same class in different catalogs, use the same class name each time. Students will not receive recommendations for classes that are named exactly the same as a class the customer has previously taken. This is another reason not to include codes, locations, or schedule information in the class name.
For example, if a student takes "Fall-18 iPhone for Beginners", they will likely receive a recommendation for "Spring-19 iPhone for Beginners" because the class names are different yet contain similar keywords. If the class name is simply "iPhone for Beginners" in both catalogs, this class will not be recommended to students who have already taken a class with the same name.
Use categories to group similar classes.
Put classes that would appeal to the same audience of customers in the same category. In general, the more specific the category name, the better.
For example, if a student previously took the "Noodles Around the World" class in the "Cooking" category, their recommendations will be more closely related to their interest in cooking than if the "Noodles" class was in a broader category called "Personal Enrichment" that also includes music and yoga classes. Depending on the program, categories for age groups (for kids' classes) or job role (for workforce training) may also be appropriate.
Use subcategories to further refine recommendations.
Classes in the same subcategory rank higher in the recommendation engine than those in the same top-level category, so use subcategories whenever possible.
Using the example above, you may have a "Personal Enrichment" category with subcategories for "Cooking," "Travel," "Music," "Language," and "Health and Wellness." Recommendations for classes in these subcategories will be much more finely-tuned to the student's past registration history than you only had one "Personal Enrichment" category that included classes with this wide variety of content. For example, a student whose registration history includes five classes in the "Cooking" category will receive prioritized recommendations for upcoming classes in that category, before they receive recommendations in the "Music" category in which they have only taken one previous class.
Use content-specific categories and subcategories.
The recommendation engine is unable to exclude categories that do not relate to the class content. Categories for times of day, days of the week, or location may reduce the relevance of recommendations.
For example, a student who has taken three dance classes in a "Weekday Mornings" category that also includes welding, conversational Spanish, and chess strategy classes is likely to get recommendations for these non-dance-related classes by virtue of this shared category. This student may miss out on recommendations for other dance classes in the "Afternoons" or "Evenings" categories if they have not yet taken a class in these categories. In this case, a Dance category would result in more relevant recommendations for students interested in dance.
Fill out instructor information in each class.
Adding instructors to all classes in your catalogs helps us recommend other classes taught by the same instructor.