Ask questions about your content
Use Adobe CX Enterprise Coworker Chat to get accurate answers about your content from Adobe Content Analytics. Ask a specific question, such as “Which images drove the most revenue last quarter?”, or a general one, such as “What kind of content should I prioritize?”. Coworker turns your question into a sound analysis.
Coworker understands how Content Analytics data is structured and applies the reporting rules that an experienced analyst would, so marketers and other business users get reliable results without deep Content Analytics expertise.
For other content performance use cases, see Content performance overview.
Best for: Business user, Analyst
Prerequisites
To ask questions about your content, you need:
- Coworker Chat connected to your Customer Journey Analytics data views. See Get started analyzing data with Coworker Chat.
- View access to the data views that you want to analyze. Coworker Chat inherits your permissions from Customer Journey Analytics.
- A data view that’s configured for Content Analytics. See Configure Content Analytics in the Customer Journey Analytics guide.
Ask a content question
-
Sign in to Coworker.
-
Select New chat.
-
In the text field, ask a question about your content. For example:
Prompt
What kind of content should I prioritize?
Coworker uses the
analyze-contentskill to answer content questions. -
(Conditional) If your question is open to more than one interpretation, answer the clarifying questions that Coworker asks, such as which content type, channel, or date range you mean.
-
Review the results.
-
(Optional) Ask follow-up questions in the same conversation to go deeper, or open a visualization in Analysis Workspace for more manual control. See Open a visualization in Analysis Workspace.
Example prompts
The following table shows example prompts for different types of content questions. Use them as a starting point and adapt them to your own content.
- “What kind of content should I prioritize?”
- “How is our content performing?”
- “What do our best-performing images have in common?”
- “Which content attributes drive the most engagement?”
- “Which images drove the most revenue last quarter?”
- “How much revenue is tied to the image at https://www.example.com/images/spring-sale.jpg?”
- “Which creative themes win in our mobile app and on our site?”
- “What imagery works for first-time visitors?”
How Coworker answers content questions
Coworker answers content questions with the analyze-content skill, which includes the following context and guardrails. The skill pairs with the cja skill and runs all queries against your Customer Journey Analytics data views through the same Customer Journey Analytics tools.
Query defaults and clarifying questions
If your question leaves out a detail, the skill applies a sensible default. If the skill is unsure what you mean, it asks a clarifying question instead of guessing.
Metric matching for assets and experiences
Content Analytics tracks two types of content: assets and experiences. Each has its own metrics. The skill always pairs the metric with the type of content that you ask about.
Content Analytics data architecture
The skill understands how Content Analytics data is structured. For example:
- Experience IDs: When you ask about a specific page, the skill filters by the Content Experience ID that identifies that page.
- AI-generated attributes: Each asset and experience is described by attributes such as photography style, colors, tone, and persuasion strategy. The skill uses these attributes to explain why content performs well, not just which content performs well.
For the full list of components, see Content Analytics components in the Customer Journey Analytics guide.
Attribution
The skill chooses the attribution model that fits your question. For example, a question about clicks on an image is answered differently from a question about how much revenue that image contributed to. For more information, see Attribution models in the Customer Journey Analytics guide.
Common analyses and workflows
The skill includes instructions for common content analyses, such as:
- Finding the attributes that top-performing assets or experiences have in common.
- Tying content exposure to conversions and revenue.
- Comparing content across web and mobile channels.
- Comparing which content works for different audiences and segments.
- Identifying content that’s overused or underused.
Visualizations
The skill selects visualizations that fit your question and the data, so that results are easy to read and share.
Tips for better results
- Start broad if you’re not sure what to ask. Coworker can turn a general question into a focused analysis, and you can refine from there.
- Say whether you mean individual assets, such as images, or whole experiences, such as pages, when you know which one you want.
- When you ask about a specific image, include its URL or asset ID. Coworker can’t identify an image by its role on the page, such as a hero image or a banner.
- Name the metric and date range you want to analyze, such as revenue last quarter.
- Specify the segments, audiences, or channels that you want compared.
- Ask about content attributes, such as imagery, themes, or messaging, rather than individual files when you want to know why content performs well.
- Add your preferred data view to Coworker Chat memory if you always analyze the same one. See Add a data view or report suite preference in Memory.
Next steps
- For more content performance use cases, see Content performance overview.
- To learn more about Content Analytics, see Content Analytics overview in the Customer Journey Analytics guide.
- For an overview of data analysis with Coworker Chat, including best practices, see Analyze data with Coworker Chat.