AI-generated themes organize related project concepts into higher-level topics. They provide a structured starting point for exploration without requiring you to define every topic in advance.
A theme is a group of related concepts representing a broader area of conversation. The generated concept lists can be reviewed, refined, and reused throughout Journey Insights Hub.
Generate themes
Use the theme-generation option at the top of the text analysis project. The themes are processed in the background and include:
A central concept list containing the identified themes
Additional concept lists for exploring each theme in greater detail
When processing is complete, refresh the project if prompted. The theme-generation control indicates that the generated themes are available.
Open the generated concept lists
Open Concepts from the side panel in Volume, Galaxy, Drivers, or Sentiment. Use the concept-list dropdown to locate the AI-generated lists.
Each list contains related concepts grouped into a broader theme.
Use the lists to:
Focus a visualization on one area of the text analysis project
Compare the reach of themes in Volume
Explore relationships within a theme in Galaxy
See whether theme concepts are associated with an outcome in Drivers
Compare sentiment across related concepts
Build dashboards around recurring topics
Review and refine the themes
AI-generated themes should be treated as a starting structure rather than a finished taxonomy. For each generated list:
Review the included concepts
Confirm that they represent the intended theme
Open Records to check how they are used
Remove concepts that introduce unrelated records
Add important language that was not included
Rename the list if a clearer business label is available
Apply colors consistently
Save the list after making changes
Duplicate a generated list before adapting it for a different audience or purpose. This preserves the original while allowing you to create a more focused version.
Create a theme from a concept search
If a specific topic is not included in the generated themes, enter it in the concept search field and select the AI option beside the search result.
The resulting list is named for the search and contains related concepts found in the text analysis project. This can help you investigate a business-specific subject without knowing all the language people used to describe it.
A search term may have no direct matches while the project still contains meaningful related language. In this case, the generated list may include concepts people used when discussing the broader topic. Review the resulting concepts and their matching records before deciding that they represent your intended definition.
What if the topic is not in the text analysis project?
If the AI does not find meaningful text analysis project language related to the requested topic, it creates a list containing a single disclaimer entry with zero matches.
The disclaimer explains that no relevant words or concepts were found in the text analysis project text. This prevents an empty result from appearing to be a valid theme.
A zero-match disclaimer means there was insufficient evidence for the topic in the project. It does not prove that the topic could never be relevant; it means the topic is not meaningfully represented in the data reviewed.
Before concluding that a topic is absent:
Check the wording and spelling of the search
Try a broader or related term
Consider how people represented in the text analysis project may describe the subject
Search the text analysis project directly for relevant concepts
Confirm that the necessary records are included
If the disclaimer remains, do not use the generated list as evidence that the text analysis project discusses the topic.
Regenerate themes
When generated themes already exist, the theme-generation option is unavailable.
Open Concepts
Locate the existing generated Themes list
Review any edits or dependencies you need to preserve
Delete the existing generated Themes list
Return to the project and use the theme-generation option again.
Deleting the existing list removes that set of generated themes. Duplicate or document any customized lists you want to retain before regenerating.
Understand the role of AI-generated Themes
AI-generated themes and Galaxy conversation clusters both help organize the project, but they serve different purposes:
Conversation clusters visualize how concepts relate within the text analysis project’s semantic map
AI-generated themes create reusable concept lists that you can edit and apply across features
Use both to explore the data from different perspectives. Clusters can reveal how conversations are structured, while theme lists provide a practical framework for ongoing analysis and reporting.