Content data

What is content data in Raptor?

Content data refers to the articles, blog posts, recipes, guides, and other editorial items you want Raptor to recommend to your visitors. Like product recommendations, content recommendations rely on tracking and a content catalog working together. This article explains the key differences between content and product data, and what you need to set up correctly for content recommendations to work.

🔍 Note: If you are working with a mixed site that has both products and content, read this article alongside Product Data: Choosing the Correct IDs for Website and Email Recommendations and Triggers.

How content differs from products

Products and content items are tracked differently because users interact with them differently. Products can be visited, added to a basket, favorited, and purchased. Content is typically only visited.

This means the recommendation engine has fewer behavioural signals to work with for content. Raptor's content algorithms are tuned to work well with visit-only data, but the quality of recommendations still depends on consistent IDs and well-structured catalog attributes.

Event	Products	Content
Visit	Yes	Yes
Basket	Yes	No
Favorite	Yes	No
Purchase	Yes	No

Because content recommendations rely entirely on visit behaviour, attributes in your content catalog (such as category, topic, and tags) carry more weight. A well-structured catalog helps the engine understand relationships between content items even when behavioural data is limited.

Content ID requirements

Every piece of content must have a unique ID. This ID is used to join tracking events to your content catalog, and it must be consistent across both.

The ID can be any stable string: a GUID, a slug, or a numeric ID
It must not change after publishing - if an article's ID changes, its tracking history is lost
Always send the content ID as ContentId in the Raptor tracking template

⚠️ Warning: Do not use a URL as your ContentId if URLs can change (for example, due to slug updates or localisation). Use a backend ID that remains stable regardless of editorial changes.

Content catalog attributes

Your content catalog is what the Recommendation Engine uses to describe and relate content items to each other. The richer and more consistent your catalog attributes are, the better the recommendations.

Recommended attributes to include in your content catalog:

ContentId - the unique identifier, must match what is sent in tracking
Title - the article or page title
Category / Topic - used for content-to-content similarity
Tags - optional, but useful for fine-grained grouping
Publication date - enables recency-based recommendation strategies
Language - required if you serve content in multiple languages
Image URL - required if content is shown in recommendation modules
Content URL - the link used in recommendation modules

🔍 Note: The more consistently attributes like category and tags are applied across your catalog, the better the Recommendation Engine can group and relate content items.

Combining content with products

Raptor can recommend content alongside products, or use content visits as an event in product recommendation strategies. This is useful for editorial-led e-commerce sites - for example, a food retailer that publishes recipes alongside shoppable ingredients.

⚠️ Combining content and product recommendations requires configuration by your Raptor implementation contact. This is not available as a self-service option.