After adding widgets, configure each one around the question the dashboard should answer. A clear dashboard uses consistent definitions, purposeful filters and enough context for another person to understand the results.
Configure each widget
Select a widget to access its available settings. Depending on the widget, you may be able to choose:
A numeric or categorical metadata field
A date field or reporting period
Concepts or concept lists
Comparison groups
A score field
Filters
Display type
Axis ranges
The available settings vary because different widgets answer different questions.
Change one setting at a time and confirm how the result changes. This makes it easier to identify which choice produced the displayed pattern.
Use widget-level filters for focused questions
Apply a widget-level filter when only one visualization should examine a particular subset. For example, you might keep the overall dashboard focused on all customer feedback while configuring one widget to show:
Low-scoring responses
One product or channel
A recent reporting period
A particular region
One customer segment
Make the filter clear in the widget title or a nearby text box. Otherwise, viewers may assume the widget represents the same population as the rest of the dashboard.
Use Global Filters for a shared subset
Use Global Filters when the same subset should apply across the dashboard.
This is useful when the complete dashboard needs to represent:
One reporting period
One product or brand
One market or region
One department
One score range
One customer or employee group
Before presenting the dashboard, confirm which widgets respond to the Global Filters and whether any widget-level settings create additional restrictions.
A filtered dashboard should clearly identify the population being shown.
Configure an AI-generated Summary breakdown
An AI-generated summary widget can summarize the text analysis project using a selected metadata field.
Select the filter icon in the upper-right corner of the widget
Open Select Summary Breakdown
Choose the metadata field to use
A breakdown can help explain differences across products, regions, sources, periods, or other meaningful groups. Before relying on the summary, confirm that the selected field supports the intended comparison and that each group contains enough records to produce a meaningful result.
Set a useful numeric axis range
Widgets with a numerical vertical axis will allow you to set a custom range. Use this when the default scale makes meaningful differences difficult to see. For example, if a score technically ranges from 0 to 10 but the observed values are all between 7 and 9, a focused range can make the variation more visible.
A narrower range can also make small differences appear more dramatic. Display or explain the selected range so the audience understands the scale.
Use consistent concepts and colors
When multiple widgets analyze the same theme:
Use the same concept definition
Reuse the same concept list
Apply colors consistently
Keep comparison groups in the same order when possible
For example, if blue represents promoters in one NPS widget, avoid using the same blue for detractors elsewhere on the dashboard. Consistent visual meaning helps the audience compare widgets without relearning the dashboard each time.
Prepare the dashboard for presentation
Open Global Filters and enable Screenshot Mode to hide the widget-building grid and other editing elements. Use Screenshot Mode when:
Capturing the dashboard for a presentation
Sharing a clean image
Presenting the dashboard during a meeting
Reviewing the final layout
Before capturing or presenting it:
Confirm all filters
Check widget titles and labels
Review axis ranges
Confirm that widgets use the intended concepts and score fields
Remove unnecessary widgets
Check that text and labels are readable
Confirm that AI-generated findings have been reviewed
Make sure the dashboard identifies the applicable period and population
Screenshot Mode changes the presentation of the dashboard; it does not change the underlying data or configuration.
Validate before sharing
A dashboard can make patterns easy to communicate, but it can also hide differences in sample size, filtering, or concept definition. Before sharing an important conclusion:
Check the number of records behind each comparison.
Confirm whether filters are global or widget-specific.
Review unexpected matches in Records.
Verify score associations in Drivers.
Compare sentiment percentages with concept volume.
Check that AI-generated text reflects the underlying evidence.
Look for exceptions or groups that the summary may have combined.