Content health

You content health score shows you how well your chatbot content covers the conversations you want to have with your chatbot users.

Your content health score doesn't worry about how accurate the chatbot was in selecting the right content, only whether the content you needed exists. If you were marking a lot of responses or catch-alls as needing to go to a new passage, this score will be a little lower.

Remember, each review depends not just on your chatbot's content, but what your chatbot users happened to send to your chatbot this time around. Don't worry if your content health fluctuates a little between reviews.

If you want to see more of a breakdown, you can switch to the Data view:

  • Expected answers indicate where the chatbot replied with the correct response.
  • Unexpected answers indicate where the chatbot should have sent an existing passage but instead replied with a different passage or a catch-all.
  • Catch-all indicates where the chatbot correctly replied with a catch-all response, such as when the chatbot user sent garbled text or a message the chatbot isn't intended to handle.
  • Missing content indicates where the chatbot didn't have the content it needed to send the correct response.

You can download a .csv file of missing content to plan the content improvements you will make to your chatbot.

Was this article helpful?
0 out of 0 found this helpful



Article is closed for comments.