Content Analytics - How it works

Hi, this is Michelle. Join me to learn about how Content Analytics works. It all starts with understanding the core building blocks, assets, experiences, and attributes. An asset refers to any image on a web page, while an experience represents all the text on that page. These elements serve as the foundation for deeper analysis. We collect event data about user interactions, such as views and clicks, with these assets and experiences. Once collected, these assets and experiences are processed using AI models to extract detailed attributes, such as color, emotion, or tone. This process, often called featurization, turns visual and textual content into structured metadata. These content engagement events are stored in experience platform, where they function just like other event data, similar to web behaviors, CRM conversions, or call center events. But now we’re adding a new type, content events. Similarly, the attributes of this content are stored as lookup datasets in experience platform. Content Analytics integrates directly into Customer Journey Analysis Workspace, making it easier to visualize, navigate, and understand how assets and experiences are performing. You’ll see thumbnails of assets, making it easier to instantly recognize which images are in question. For experiences, the system takes upper page screenshots so that content creators get immediate visual context. The detailed assets view panel allows you to inspect key metadata, including impressions, the number of experiences it has appeared in, and the first and last impression date, and the AI-derived attributes, such as themes, visual elements, tone, and contextual labels. Content Analytics provides a pre-configured reporting template available out of the box. This template tracks standard metrics for both assets and experiences. Users can slice and dice this data however they need, by asset type, page location, audience segment, or campaign, and apply custom filters and calculated metrics, just like in any CJA reporting scenario. Let’s take a deeper dive into how Content Analytics actually works under the hood, specifically the data flow and infrastructure that powers it. We’ll start with the baseline. In today’s CJA environment, web activity is captured through standard mechanisms, either via their web SDK or through data connectors like the analytics source connector. These events are processed by the experience platform and flow into CJA’s back-end reporting engine. With Content Analytics, an additional content engagement event is recorded. These entail asset views, experience views, asset clicks, and experience clicks. They represent precise content-level interactions captured exclusively from the web SDK. Here’s how it works. When a user views or interacts with the tracked asset or experience on a web page, the web SDK captures that event, particularly the URL of any asset or page. These events are sent into the experience platform as separate event streams, distinct from a standard page view or click events. They’re stored in their own data set, using a standardized schema designed for Content Analytics. Once the experience platform receives a new event, it triggers an AI-driven processing flow. First, the system fetches the actual image or experience content from the URL. Next, it passes that content through an AI identity and attribute service. This AI identity service performs two critical tasks. First, identity resolution. Has this image been seen before across your digital ecosystem? If so, it reuses an existing ID. If not, it generates a new unique identifier. Second, attribute extraction. For next assets and experience, the service analyzes the content and assigns metadata attributes, such as outdoors, winter scene, or blue color tone. These derived attributes are then written into lookup tables, one for assets, one for experiences, and stored alongside all other experience platform data. From here, the content events, metadata, and attributes merge with your broader web data inside the CJA reporting engine. Now, asset and experience level data sit side by side with traditional behavioral data like page views, product views, or purchases. You can now run reports that correlate content level interactions with conversion events, engagement funnels, and business KPIs. And because this architecture shares the same identity infrastructure, the event correlation is seamless. A content view and a purchase from the same session tied together. An asset click that leads to a download days later tracked across the entire journey. This entire pipeline brings a new dimension of intelligence to content, allowing marketers and content teams to understand not just what content exists, but how it performs. Thank you for watching this video about how content analytics works.