Aplysia 3 Report

Insights & Metrics

The Aplysia 3 Insights report allows clients to measure performance of the chatbots of type Aplysia 3:

Key metrics cards include:

  • Automation Rate (conversations with no staff intervention)

  • Chatbot CSAT (customer satisfaction score with the performance of the chatbot)

  • Unanswered Questions (number of questions that were not answered by the chatbot automatically)

  • Negative Feedback (number of answers with a negative feedback, i.e. tumbs-down on answer)

Graphs include:

  • Answered questions versus Unanswered questions (number of question that were answered by the chatbot vs number of questions the chatbot could not answer)

  • Number of answers generated by the chatbot (and how they were generated) versus staff intervention (conversations transferred to the console)

    • Chatbot answers:

      • Topic answer: an answer was provided with the content of the topic only (no access to the document)

      • Complementary answer: an answer was created by using content from the document, with buttons and/or images from the topic content

      • Document answer: an answer was created by using the document only

    • Staff intervention:

      • Contact Staff Form: the chatbot could not create an automatic answer for the user and the conversation was transferred to an agent

 

Tables include:

  • Messages with negative feedback from the users

    • In this table, the clients can see details about the automatic replies that had a negative feedback from the user (tumbs down):

      • Message: what the message from the user was

      • Answer: what the automatic message from the chatbot was

      • Actions: actions that can be taken by the client to improve the answer provided by the chatbot

        • Improve topic: the client will be redirected to the content of the topic to improve it

        • Improve document: the client will be redirect to the document to improve the content

  • Messages that the chatbot was unable to provide an automatic answer

    • In this table, the clients can see the messages that did not receive an automatic reply from the chatbot:

      • Message: what the message from the user was

      • Property context: from what property the message was related to

      • Actions: actions that can be taken by the client to add content, so the user can receive an automatic reply.

        • Improve topic: the client will be redirected to the content of the topic to adjust it

        • Improve document: the client will be redirect to the document to add the content