Analyzing how your bots perform is essential to understand the behavior of your bot. You can learn more about how your bots behave, get insights into what parts of the stories work and what interactions require additional improvements by exporting your bot data and analyzing it. With the power of exporting your data, you can determine which bots have the highest conversation completion rate and also a breakdown of how your visitors interact with each bot action.
A quick guide to exporting your data:
- Login to your Freddy Self-service account as an Admin or Owner and click on Settings > Export report
- From the dropdown, pick the bot and the bot’s version for which you want to export the data
- You can choose to export four different types of reports:
- Customer Information and Contextual data - The report fetches information about your customers (name, email, phone number, created date, etc.)
- Customer conversations - This contains all the conversations initiated by the customers (including the preview).
- Dialog count - This provides a list of all the dialogs that the customer traversed along with the count.
- API count - This gives the number of times the bot hit the APIs that you had configured
- You can choose between
- exporting the report once (One time report), where you pick a time range and get the export immediately, or
- export the report regularly (Schedule export), where you can choose between different frequencies
- The export of the entire data, in CSV format, will be sent to an email address that you enter
Note: If you’d like to export your bot data to AWS instead of receiving it via email, we can help you do so by pushing all your data to S3. Learn more here.
Customer information and contextual data report fields
- Ticket Hash - A unique ticket identifier for every conversation that users have with the bot. You can map this ticket hash with the ticket hash in the other file.
- Conversation Hash - One conversation can have multiple tickets. For example, if the customer clicked on start over, a new ticket is created, but the conversation remains the same.
- Bot Name - The name of the bot where the conversation took place.
- Bot Version Number - The bot version where the conversation took place.
- Node Group - The last Node Group executed in this ticket.
- Channel - The source through which the chat with the bot is initiated. The channel can be a website, Facebook account, etc.
- Customer Name - The customer’s name, as shared in the chat or passed in the session while initiating the bot
- Customer Email - The customer’s email, as shared in the chat or passed in the session while initiating the bot
- Mobile - The mobile number of the customer provided while in chat or if
- passed in the session while initiating the bot
- City - The city from where the customer initiated the bot. This information will come via the IP
- State - The state where the customer initiated the chat. This information will come via the IP address.
- Landing URL - The page of the website on which the customer first landed.
- UTM Source - If any UTM source is mentioned in the landing URL
- UTM Campaign - If any UTM campaign is mentioned in the landing URL
- Ticket Status - It shows if the status is open (the conversation is ongoing),
- closed (conversation ended) and temporary/new (if the customer has not initiated the conversation yet).
- Extra Info - Any extra information stored in the ‘tckt:xtrInfrmtn’ variable using the update database or any information passed in the ‘getClientParams’ function while initiating the script.
- External Ticket Id - The external ticket id passed in the ‘xtrnlTcktId’ variable in
- ‘getClientParams’ function while initiating the script.
- Create Date - The date at which the ticket was created.
Customer conversations report fields
Dialog count report fields
- Dialogue ID - Gives you the id of the dialogues that were invoked during the flow.
- Count - Number of times a dialogue was called.
API count report fields
- API ID - Every API that is configured has an ID to it. This one has the list of IDs that were called during the flow.
- Count - Number of times an API was called
Please write to email@example.com if you have any more questions; we’ll be happy to help you.