Setting up Response Metrics (Survey-Level Custom Metric)
Updated
You can now create response‑level derived metrics directly within the CFM survey, giving you the ability to compute new values based on existing survey question metrics without the need for exports or offline processing. These metrics are applied only to responses received after the metric is created, while historical responses collected earlier can be backfilled on request, ensuring flexibility and accuracy in your reporting.
Business Use Cases
Response metrics help to combine multiple question scores (e.g., CSAT Q1 + CSAT Q2) into a single metric.
Response metrics build weighted KPIs such as a weighted CSAT or composite satisfaction index.
Response metrics help to normalise and transform response data for consistent use in dashboards.
Enable response-level derived measures to appear as columns in the Response tab for micro-analysis.
This capability removes dependency on exports and offline calculation, allowing you to perform analysis directly within CFM. It enables richer analysis by providing advanced insights without external tools, while ensuring that metrics are reusable across dashboards for consistency and efficiency. Additionally, it offers null‑safe handling, automatically excluding Not Answered and N/A values from derived metrics to maintain accuracy and reliability in reporting.
Prerequisites
You would need Edit Survey Custom Fields access, along with Survey Settings to configure and manage custom fields within surveys.
View Survey Dimension: You would require access to View Survey Dimensions and use survey-specific fields (question-level fields) when creating response metrics and applying them in dashboards.
View Response and Analytics: You would require access to Response and Analytics to view dashboard data that renders response metrics and supports analysis.
Setting Up Response Metrics
Go to Settings and navigate to Response Metrics under Survey Custom Fields.
Click + Response Metric.
Go to Create Response Metric and fill in the details:
Metric Name: You can add a meaningful name to the metric.
Description: You can add a meaningful description to the metric.
Metric Formula: You can build a formula using:
Survey question metrics only (e.g., CSAT, Rating, Slider, Matrix, Rating Matrix, All Statement Rating).
Other Response Metrics created for the same survey can also be used to create complex nested response metrics as per use case.
Arithmetic operators: +, -, *, /.
AVERAGE(...) function to calculate averages across selected metrics. AVERAGE ignores “Not Answered” and “N/A” values per response.
Click Save.
Allowed metrics (can be used as inputs in a Response Metric):
Question Type | Metric |
Text |
|
| Insights Count |
| Experience Score |
NPS |
|
| Rating |
CSAT |
|
| Rating |
CSAT Multi Statement |
|
| Rating |
| All Statement Rating |
Slider |
|
| Rating |
Slider Multi Statement |
|
| All Statement Rating |
Rating Matrix |
|
| All Statement Rating |
|
|
| Response Quality Score |
Other Response Metrics created for the same survey, can also be used to create nested response metrics. |
|
Note: Every Response Metric undergoes assessment and storage upon the arrival of a response. Values are filled in exclusively for responses received subsequent to the metric's storage.
Null Handling
If every value within the formula is "Not Answered"@1, the resulting response metric will also be "Not Answered". This is represented as "-" in both the dashboard and the response tab.
When certain inputs lack responses, the metric relies on the data it does have for each response. The Not Answered values are not taken into account when widgets aggregate data.
Processing window
Response metrics are calculated solely for responses collected after the metric has been saved. Previous responses predating the metric's creation will not include the derived metric value and will be categorized as "Not Answered" for that particular metric.
Turning off a metric will prevent it from being calculated in the future, but any previously stored values will still appear in responses where they were already present.
Example: Creating an average of NPS + CSAT
First normalize CSAT (out of 5) to a scale of 10 by creating a weighted CSAT metric: CSAT * 2.
Next, create an AVERAGE Response Metric and select NPS and Weighted CSAT.
Alternatively, create a single AVERAGE Response Metric with the formula: AVERAGE(NPS, CSAT * 2).
How to use it?
Response tab:
Once a response metric is created for a survey, you can add the response metric to the response tab as a column.
You can access the “Add column to Right/Left” option from the response tab columns.
You can select Response Metrics created for the survey from the “Response Metrics” bucket present in the selector dropdown.
Custom Dashboards:
Response Metrics appear under a new bucket in the widget builder: Go to User Defined Metrics and naviagte to Response Metrics.
Similar to standard dimension and measurements created using survey questions, response metrics are also tagged with the name of each survey. Thus, we can also be filtered by survey name in the widget builder.
Plotting a Gauge Chart using the Average NPS + CSAT.
To validate response metrics, navigate to Survey Settings and go to Response Metrics and create a simple formula such as (CSATQ1+CSATQ2)/2. Confirm that the metric appears as a new column in the Response tab, then add it to a custom dashboard through the Response Metrics bucket. Ensure calculations correctly handle Not Answered values, and when cloning the metric, verify that the formula remains locked while only the Name and Description fields can be edited.
Key points to note
Response Metrics are computed only for responses received after the metric is created. Historical responses are not backfilled.
Formula edits are not allowed once a metric is in use. Only Name and Description can be edited.
Response Metrics are survey‑specific and cannot combine data across multiple surveys.
Response Metrics cannot be used as filters or drilldown dimensions.
Certain measurement types are incompatible: NPS score, CSAT scored, Rank Order measurements.
Numeric Profile fields, transaction fields, and external dataset measures cannot be used as inputs for Response Metrics.
Manage Response Metric
Enable metric: You have the option to enable the metric. If you choose to deactivate a metric, it will prevent any further calculations from being made. However, any previously stored values will still be included in responses where they were already present.
Filters: You can filter using enabled and created by filters.
Access the Vertical Ellipsis against a response metric and you will be able to edit and clone the metric.
Edit: Helps to edit only Name and Description are editable.
Clone: Helps to create a copy with the same formula (the formula cannot be edited once created).
Best Practices
Use descriptive naming (e.g., Weighted CSAT Checkout).
Document weightings or logic in the description.
Validate manually for accuracy of logic formula before wider use.