Dimensions Analysis
A plain CSAT or NPS report gives you the overall number. Dimensions Analysis tells you how that number differs across segments — and which segment values correlate with your promoters versus your detractors. Pick a survey, pick a dimension (a survey answer or a context attribute like device, browser, store, "would you buy again? yes/no", or "could you order clearly from X? yes/no"), and the report recomputes every score for each value of that dimension.
If you're a CX analyst or product owner, this is the report that answers "does satisfaction differ by device, by whether they'll buy again, by order clarity?" It's the cross-tab engine of the family: same metric, cut by the answers and metadata you choose, with a strong correlation story built in.
What you'll see
A typical Dimensions Analysis report includes:
- NPS and CSAT gauges — the survey's overall scores, then the same KPIs recomputed for each value of the dimension you've chosen.
- A response-distribution donut — how responses split across the dimension's values.
- "NPS by [dimension]" and "CSAT by [dimension]" charts — bar-plus-line views showing the score for each value with response counts overlaid.
- Responses, percentage and delta tables — the numbers behind each segment, versus the prior period.
- A monthly stacked bar — how the dimension's value distribution shifts over time.
- A grouped bar of rating-metric scores by dimension value — see which attributes drive the gap between segments.
Most of these sections repeat per dimension value, so each segment gets the full treatment.
The defining control is the dimension selector — it sets the cut for the entire report. Alongside it, filter by workspace, questionnaire and date range, and use the include-respondents toggle to bring respondent-level detail in or out.
Put it to work
- Find the segment gap. Compare the per-value gauges to see where satisfaction splits — for example, mobile NPS far above desktop.
- Tell the correlation story. Look at where detractors cluster — say, where "order clearly?" = No — and you've found a concrete, fixable root cause.
- Watch the mix move. The monthly stacked bar shows whether a segment is growing, so you know whether a low-scoring value is a rising risk.
- Pinpoint the driver attribute. The grouped rating-metric bars reveal which attribute pulls one segment below another.
Pick a dimension with a decision attached
The most useful cuts are the ones tied to an action you can take — device (a UX fix), order clarity (a content fix), store (an operational fix). Start from the question you want to answer, then choose the dimension that splits it cleanly.