How Does CSAT Feedback Flow Work?

There is a flow in which this mechanism works, and it starts with the bot asking the user, “Was I able to help you?”, and the answer to this would be either Yes or No.

If the user messages No, we take feedback as 1. The bot was not able to help so the rating is 1. Further, the bot tries to take qualitative feedback (comments) from the user, as shown

Note

CSAT cares only about the percentage of Satisfied (Yes) ratings.
After a user has clicked on “No“ as an answer to “Was I able to help you?“ it is evident that they are not satisfied with the bot’s responses, so rather than annoying the user further by asking to rate on a scale of 1-5, if they have said No, we take it as a “non-satisfied user”, and the rating could either be 1,2 or 3, but default 1 makes the most sense logically.
Ideally, for a bot, if the user asked a question, to which the bot gave an answer and then for the bot’s question “Was I able to help you?” the user’s reply was “No”, then for the bot, it has failed which is why we take the rating as 1.

If the user messages Yes, the bot would ask the user That’s good to know. On a scale of 1 to 5, how would you rate the overall experience?

For every rating that the user gives, there is a response available with the bot, so as to maintain the flow for collecting the feedback.