Subtopics

The Subtopics page provides a clear view of what users talk about the most when interacting with Smart Assistant. It helps you understand the main areas of interest, frequent questions, and the intent behind user conversations, allowing you to measure how well your assistant covers the most relevant subjects.

By reviewing the data on this page, marketing and operations teams can identify trends, adjust content or automations, and ensure that the assistant is aligned with customer expectations and product priorities.

Period selection and filters #

At the top of the page, you can select the Time Range you want to analyze. The available options are Last 24 hours, Last week, Last month, Last year, or a Custom range.

You can also use the Filter panel to refine your analysis by the following parameters:

• Escalated : show only conversations that were transferred to a human agent or support team.
• Rated : filter conversations that received a satisfaction rating from users.
• Source : filter results based on the user’s device or environment (desktop, mobile, tablet, app, etc.).
• Channel : filter by the communication channel (website, WhatsApp, Messenger, etc.).
• Language : focus on a specific language used in conversations.
• Country : limit the data to a particular country or market.

Filters can be combined to reveal insights such as “Which subtopics generate escalations from mobile users in France?” or “Which subjects get the highest satisfaction ratings?”

Subtopics table #

The main table lists the Subtopics detected by Smart Assistant during the selected period. Each row represents a specific theme or intent extracted from conversations, such as Orders, Returns, or Payments.

Columns explained #

• Subtopic :: the name of the detected subject or intent, automatically categorized by Smart Assistant’s classification engine.
• Classifications : number of times the assistant identified this subtopic in user conversations. This indicates the subtopic’s frequency and relative importance.
• Distribution : shows the percentage share of each subtopic among all conversations, helping you understand which subtopics dominate user interactions.
• Conversations : number of complete Conversations associated with the subtopic.
• Usage :- percentage of all assistant interactions that included at least one message classified under that subtopic.
• Average Rating : average user satisfaction rating (if available) for conversations related to that subtopic.
• Feedback Rate : proportion of users who provided feedback compared to total conversations in that subtopic.

These columns allow you to identify which subtopics generate the highest engagement or require optimization — for instance, if Returns represent a large share of interactions but have low satisfaction, that may indicate the need to refine your responses or escalation flow.

How to interpret subtopic data #

Understanding subtopic distribution helps you assess how effectively your Smart Assistant handles real customer intents.

• High-frequency subtopics like Orders or Delivery show recurring interests or operational dependencies.
• Low-frequency or emerging subtopics may reveal new user needs, product interest, or upcoming issues to monitor.
• A balanced distribution of subtopics typically reflects a mature and diversified assistant, whereas concentration around a few subtopics can signal opportunities for content or feature expansion.

Combining these insights with filters such as Rated or Escalated provides a deeper understanding of user satisfaction and automation coverage.