Skip to main content
Chatbot CSAT

Understand customer satisfaction with your chatbot (Customer Facing Workflow) conversations by sending a CSAT survey.

Brian Donohue avatar
Written by Brian Donohue
Updated over 3 months ago

You can understand how well your chatbot (Customer Facing Workflows) conversations are performing with a chatbot CSAT survey. You'll be able to measure your customers’ satisfaction with every chatbot conversation, and identify conversations and areas that are performing well or in need of improvement.

Chatbot CSAT is only available with Workflows (available on some plans).


Add chatbot CSAT to a Workflow

By adding the Ask for conversation rating step to your chatbots (Customer Facing Workflows), a survey can appear in a conversation when:

  • A customer completes the CSAT step in a chatbot Workflow

First, you'll need to create a new path in a new or existing Workflow, click the Add step button and select the Ask for conversation rating step.

A CSAT survey will be sent to the customer if their interaction proceeds to the Path in the Workflow where the Ask for conversation rating step has been added.

Prevent customers from rating after a given period of time

Click on the CSAT widget if you want to "Prevent customers from rating after a given period of time" and then specify the days/hours/minutes.

Note:

  • Customers will still be able to see and interact with the CSAT survey in the Messenger, but their response won't be recorded after this period of time.

  • This setting can also be configured in Basics which will be used as a fallback if there is no value set in the CSAT widget.

Prevent customers from changing their rating after a given period of time

Click on the CSAT widget if you want to "Prevent customers from changing their rating after a given period of time" and then specify the hours.

Note:

  • This setting will not prevent customers from submitting an initial rating, it will only prevent them from changing their rating. Use the "Prevent customers from rating after a given period of time" toggle instead if you don't want customers to leave any rating after the specified time.

Wait for customers to give a rating before continuing the Workflow

Click on the CSAT widget and toggle on “Wait for customer to give a rating before continuing the Workflow”.

This is helpful if you'd like to take specific actions in the Workflow based on what the rating was, using conditional branches (see example below).

When are CSAT surveys sent?

CSAT surveys are only sent when a conversation contains a ‘back and forth’ interaction.

A ‘back and forth’ interaction means there has to be at least one message from a customer and also at least one message from a teammate or bot, this can be Fin AI Agent or a chatbot (Customer Facing Workflow).

Other things which may affect a CSAT sending:

  • If a conversation contains multiple customers a CSAT survey will not be sent. This is because it’s not possible for multiple people to rate the same conversation.

  • You can turn on a general CSAT rule for your customers in the Basics setting section. This rule will be superseded by any other CSAT workflow you create.

When multiple agents are involved, who gets the CSAT rating?

When a chatbot (Customer Facing Workflow) sends a CSAT rating survey in a conversation with multiple agents, the rating attribution follows this order:

1. Teammate

2. Fin AI Agent

3. Chatbot

Only the most recent visible agent in the conversation will be rated. If no agent is visible, we check for a Fin AI Agent and then chatbot. The Workflow will only send one CSAT survey per agent.

Who gets the CSAT rating if Fin reached the resolution after teammate already participated in the conversation?

If Fin AI Agent participates and resolves a conversation after a human teammate has already participated, CSAT survey will be sent for the last human teammate who participated in the conversation, rather than for Fin.

For example, if a teammate passes the conversation to a reusable workflow to let Fin answer and Fin CSAT is turned on, if Fin resolves the conversation the CSAT survey will be sent for the teammate.


Chatbot CSAT reporting

You can only report on chatbot CSAT by adding the chatbot metrics and attributes to charts in custom reports.

The chatbot metrics can be viewed by or segmented by the attributes in charts, and are all a part of the Conversation Rating dataset.

There are 5 chatbot metrics available:

  • Chatbot CSAT score - Percentage of positive (😀 or 🤩) conversations ratings for chatbots out of all conversations ratings for chatbots.

  • Chatbot DSAT score - Percentage of negative (😠, 🙁 or 😐) conversations ratings for chatbots out of all conversations ratings for chatbots.

  • Chatbot positive conversation ratings - Number of positive (😀 or 🤩) conversation ratings for chatbots.

  • Chatbot negative conversation ratings - Number of negative (😠, 🙁 or 😐) conversation ratings for chatbots.

  • Chatbot conversation ratings - Number of conversation ratings for chatbots.

There are 3 Chatbot Attributes:

  • Chatbot rated - This filters by the rated chatbot involved in the conversation. The name of the Customer Facing Workflow.

  • Conversation rating - This filters by the conversation rating.

  • Conversation rating remark - The comment left by a customer after submitting their conversation rating.

Chatbot CSAT is currently excluded from the Overview report.

You can also + Add filter and select “Chatbot rated”, which allows you to filter the report by your Chatbot Customer Facing Workflows.

To view Remarks from customers you can use the Chart Drill-in feature on either the Chatbot CSAT or DSAT score, the Chatbot conversation ratings, Chatbot positive conversation ratings or Chatbot negative conversation ratings metrics.


Chatbot CSAT viewed via the Help Desk

The chatbot CSAT rating will appear in the conversation. This way, Teammates have upfront context on how a customer is feeling before they engage with them.


💡Tip

Need more help? Get support from our Community Forum
Find answers and get help from Intercom Support and Community Experts


Did this answer your question?