Understanding the differences between Standard Custom Themes and Deep Themes will help you capture customer feedback effectively based on your specific tracking goals and use cases.
Overview of Theme Types
Our platform offers two distinct theme types to help you analyse customer verbatims. While they serve different purposes, both are available side by side and can be applied independently to the same datasets across Explore, Heartbeat, and AcceleratorAAI (AAI) reports.
Note: A single custom theme cannot currently contain a mixture of standard and deep theme groups.
What are Standard Custom Themes?
Standard Custom Themes rely on specific keywords, phrases, or predefined topics to categorize customer feedback. They look for exact language matches within your dataset.
Benefits & Use Cases
Precise & Predictable: Standard themes are ideal when you need exact match precision against known terms.
Established Term Lists: Perfect when you already have a defined list of keywords, brand names, or specific language you want to track.
Compliance & Regulation: Essential for identifying specific, mandatory wording for legal or regulatory requirements.
What are Deep Themes?
Deep Themes are a new type of custom theme powered by vector search. Instead of matching exact keywords or phrases, they find verbatims that are similar in meaning to a description you provide in everyday language.
This allows you to describe the core feedback you want to capture without having to anticipate every possible word or phrase a customer might use.
Benefits & Use Cases
Semantic Matching: Deep themes look for context and meaning rather than strict word matches, making them ideal for broader concepts expressed in varied ways.
Flexibility Over Time: Standard themes can miss feedback when customer language evolves. For example, if a food and beverage theme tracks the word "Coca-Cola" and a stadium switches to serving Pepsi, a standard theme might miss those new references. A Deep Theme recognizes the underlying intent and continues capturing feedback about soft drinks, even when specific brands or wording change.
Standard vs. Deep Themes: Summary Comparison
Feature | Standard Custom Themes | Deep Themes |
Search Mechanism | Keyword, phrase, and topic matching | Vector search (semantic similarity) |
Setup Requirements | Predefined list of specific terms | Natural language description of intent |
Best For | Exact matches, specific terminology, regulatory tracking | Capturing broad concepts, evolving customer language |
Flexibility | Rigid to specific phrasing | Highly flexible to changing words/brands |
Using Standard and Deep Themes Together
You don't have to choose just one approach for your entire workspace. Yes, both theme types can be used together. They operate alongside each other in the platform and can be applied independently to the same datasets to give you both exact-match precision and conceptual depth.