Customer Satisfaction Score, or CSAT score, is an often used metric in the customer experience space. Typically, it is presented to customers as a survey, with a sliding scale from 0 to 100 that is meant to measure customer happiness with their support interactions.
CSAT is one of the main metrics that can provide insight into how your customers feel about your product, service, or support experience and is incredibly important to help support leaders track the performance of their organization.
In Forethought's widget, we support a variation of CSAT that can be used to understand, optimize, and measure results.
How to Activate
To activate the CSAT survey for your widget, you will need to reach out to your Forethought Implementation Team. After understanding more about the information you would like to collect, they will be able to enable your CSAT survey.
Once enabled, the CSAT survey is displayed to customers when the conversation has been through a minimum number of steps (which we are able to adjust is desired) AND
When the user has closed the chat by clicking on the top right corner.
The user has gotten to the end of a workflow with no more steps remaining.
The Survey that Customers Receive
The information collected during the CSAT survey is standardized for data collection and cannot be changed. It is comprised of:
- A numerical selection of 1 to 5
- Categorical and qualitative feedback with different selections based on the numerical response.
- Categorical selection of "Was your question resolved?" (Yes/No).
By standardizing these inputs, it makes it easier for admins to see trends in the data, associated with various paths and intents within the widget.
How to Interpret the Results
Based on the initial numerical selection, the CSAT survey will display different options to provide one or more categories of feedback. The various feedback selections display a range of issues or positive experiences that can be used to optimize the intents or workflows.
The buttons provided follow the below logic:
|Took too long
|Did not understand
|Understood my needs
For some of the more frequently selected problem categories, the selections typically translate to:
- “Did not understand” → Set up more intents or improve intent phrases
- “Unhelpful response” → Inspect and change workflows, check knowledge
- “Impersonal” → Make Context-Variable driven workflow personalization
- “Unfriendly” → Change language/entry step and ask about humanizing through SupportGPT
- “Took too long” → Set up a more direct path and workflows to support
Digging into the Root Cause
When seeking to interpret CSAT scores further, the best practice is to follow this sequence of steps:
- From the top-level view, identify the most frequently selected category (such as Unhelpful or Quick Turnaround).
- In the table, the categories should be grouped by intent. By looking at how the qualitative feedback is attributed to different intents, find an intent worth inspecting further.
- Once an intent is identified, navigate to that intent in the Workflow Builder main canvas page.
- Starting from the top down, zoom into each sub-branch with high traffic and low CSAT. This can be identified via the "Metrics" button that exists at each step of the workflow.
- Keep drilling down until you land on a text step/action/article that appears to be the root of the low CSAT and make any required changes.