Use Ask Whatfix AI

Use the following steps to access Ask Whatfix AI:

1. On the Whatfix Analytics dashboard, click Ask Whatfix AI.

   

2. Enter the question in the text box such as “What is the overall Whatfix Content Health and usage”
   

3. Click the Ask icon to submit your question.

   

   Info:

   • If you’re unsure of what questions to ask, click a tile containing a predefined question.

     

   • Once you submit the prompt, a Trend Insight, Funnel Insight, or User Journey chart appears to answer your question. Ask Whatfix AI shows the events and metrics used to generate the various charts.

   • Change the date filters, metrics, and chart types as required for Trend Insights.
     

   • In Funnel Insights, change the date filters and compare the data for a custom date range.
     

   • In User Journey, change the date filters and the Journey steps shown (number of levels).

     

4. Click Open in new page to drill down on the data with more filters.

   

   Note:

   An editable Trend Insight, Funnel Insight, or User Journey is automatically generated for your query. Whatfix enables you to customize the Insights by adding filters or breakdowns, and then save the Insight.

   • Trend Insight:

     

   • Funnel Insight:
     

   • User Journey:

     

5. Ask any follow-up questions, if required.

   

Info:

Access Ask Whatfix AI in the following ways:

• Click Ask from the Dashboards or Insights section on the Whatfix Analytics dashboard.

  

• Click Create insights from the Insights section on the Whatfix Analytics dashboard.

  

• Click Ask from the Guidance analytics on the Whatfix Guidance dashboard.
  

Use @ to mention a specific event in your Ask Whatfix AI prompt. This helps Ask Whatfix AI understand exactly which event you want to analyze, instead of relying on the event name in your question to determine the intended event.

Note:

When you type @, you can search for an event by name. Hover over an event to view additional details, such as its description, when it went live, last interaction, and the Insights that use it.

For Example:
To understand user engagement with a specific User Action, such as Claim Details, mention the user action using @, then type “How many times did users click @Claim Details last month?” as your question. Ask Whatfix AI uses the selected user action as context to generate insights specifically for Claim Details.

After you submit a prompt, Ask Whatfix AI analyzes the available engagement data and generates insights. It reasons through the problem, explains its approach, breaks the analysis into logical steps, surfaces multiple insights, and suggests next steps. Depending on the query, the response can include charts, tables, trends, comparisons, or recommendations.

• Show thinking: Displays the analysis steps performed while generating the response. It analyzes your question gathers the relevant events, filters and breakdown content, and generates editable charts.

• Generate Insights: After you submit your prompt, Ask Whatfix Al generates a detailed analysis with key metrics, editable charts, tables, breakdowns, recommendations, and key observations from the analyzed data. It can generate:

  • Trend Insight analysis to track how often events occur, how many users perform actions, and how metrics change over time (daily, weekly, or monthly).

  • Funnel Insight analysis to understand step-by-step user conversion and drop-off.

  • User Journey analysis to reveal what users do before or after a key event.

  For Example:
  For the prompt “What is the overall Whatfix Inhalt Health and usage?”, Ask Whatfix AI displays the analysis steps it performs and then generates a report with a content health summary, key metrics, engagement trends, visualizations, recommendations, and suggested follow up questions.

  

• You can also ask product knowledge based questions such as explain Whatfix features (Flows, Smart Tips, Beacons, Popups, Self Help, Task Lists, and more) or how to create or configure Whatfix content.

  For Example:
  For the prompt “What is Self Help and what are the associated analytics events?”, Ask Whatfix AI explains the Self Help feature, lists the associated analytics events, and provides links to the related documentation.

  

Ask Whatfix AI stores all questions from your current session under Chats.

Use the following features to manage your chat sessions:

• Identify sessions: Whatfix names each chat using the first question you ask in that session.

• Access previous sessions: Open old chats at any time to load all questions asked in that session.

• Continue the conversation: Ask new or follow-up questions within the same chat.

• View recent activity: The last five chats appear under Recent.

  

• Review older chats: Access other chats under History.

  

• Organize or remove sessions: Hover your cursor over the chat, and then click the More icon to rename or delete a chat.

  

  Info:

  • Hover your cursor over a chat to view the entire chat title. Rename if required.

    

  • To start a fresh session, click New chat.
    

  • All chat sessions remain in your history for one year from the date they are created.

When Ask Whatfix AI detects ambiguity in a query, it prompts you with clarifying questions to better understand your intent before generating an Insight. This can occur when multiple events, cohorts, user lists, or date ranges match the query. Select the most relevant option from the suggested cards.

For example, if you ask, “How many users from ES came to the app in the last 20 days?”, Ask Whatfix AI may identify multiple matching events and prompt you to choose the relevant event.

After you submit a prompt, Ask Whatfix AI suggests up to three follow-up questions to help you explore the data further. These suggestions are based on the current Insight and can help you analyze trends, apply breakdowns, compare time periods, or focus on specific user segments without having to write a new query.

For example, after asking “How many users interacted with Whatfix content in the last 30 days?”, Ask Whatfix AI might suggest follow-up questions such as:

• “How has Whatfix content engagement trended over the last 30 days?”

• “Which type of Whatfix content (Flows, Popups, Beacons, etc.) had the most user interactions?”

• “Which user segments or departments had the highest Whatfix content engagement?”

Click a suggested question to continue the analysis and generate a new Insight based on the current context.

The following are some sample queries that you can try out:

• How many times users completed User Actions in a particular month?

• How many users clicked the Sign up button?

• How many Flows are completed?

• How many users clicked a link from Self Help?

• How many users skipped a Pop-up?

• How many times was a Flow completed breakdown by trigger location and segment name?

• How many users opened Self Help breakdown by country?

• How many users dropped off from a Flow?

• How many users are completing the Onboarding process?

• How long are users taking to complete a process?

• How much time are users spending on a process?

• What are the top paths users are taking from a specific page?

• What are the top paths users are taking from a specific button or User action?

• What are the different paths users are taking to reach a specific button, page, or User action?

• Where are users dropping off in the lead creation journey?

• What are the most common paths users are taking from a specific button?

• What are users doing on the application after clicking a specific button?

• How are users reaching a specific button or page?

• Are there any deviations from the ideal path from a specific button?

• Where are users navigating to after clicking a specific button?

• What are the users from the sales department doing after clicking on a specific button?

• How are users from India reaching a specific page or button?

• The following are some sample queries that you can try out:

• How can I improve adoption of a specific Flow across my user base?

• Where are users dropping off in a specific process, and how can I reduce it?

• What can I do to boost engagement with Self Help content this quarter?

• Which user cohorts need the most help to start using a specific capability?

• How can I shorten the time it takes to finish a specific Flow?

• Which navigation paths most often lead to successful use of a specific feature, and how can I promote them?

• How can I increase the completion rate of a specific Flow?

• What actions will drive more repeat usage of Smart Tips?

• How do I improve search-to-success conversions in Self Help?

• Which content types should be expanded to increase overall Whatfix engagement?

• What interventions will help users finish a specific Flow?

• Which steps in a specific workflow cause the most friction?

• How can I encourage more users to explore specific features?

• What guidance can steer first-time visitors quickly to a specific feature?

• How can I improve adoption of Whatfix content on Mobile compared to Web?

• Which user journeys lead to the highest engagement with a specific feature, and how can I replicate them?

• How can I accelerate task completion for a specific process?

• What can I do to convert more Beacon hovers into clicks?

• How can I improve the helpfulness of the Self Help search results?

• How can I increase engagement with Pop-ups?

• What can be done to improve application adoption among users in a specific department?

• How can I encourage power users to try a specific feature?

• How can I reduce unnecessary detours before users reach a specific feature?

• How can I improve conversion through a specific funnel?

• Which steps in a specific Flow need optimization to lower exits?

• How do I speed up a specific process?

• Which user segments struggle with a specific feature, and why?

• Where should I intervene to reduce drop-offs in a specific workflow?

• How do I optimize the funnel leading users to access a specific feature?

• How can I encourage more users to finish a specific Flow?

• How can I increase feature discovery for a specific feature?

• How do I ensure users find and adopt a specific feature faster?