Cross Tabulation

How to cross two variables to compare segments in kicue, including row % vs column % and using URL parameters as axes.

Cross tabulation combines two questions to reveal segment-level differences. Think "satisfaction by gender" or "brand recall by age bracket" — cross tabs are essential for comparative analysis.

Setting axes

From the Analytics tab, pick Cross tabulation to open the axis configuration screen.

  • Row axis (banner) — the attribute you want to compare (gender, age, region, ...)
  • Column axis (stub) — the question you want to summarize (satisfaction, purchase intent, ...)

The table updates instantly as you adjust axes.

Reading a cross tab

Example: gender × satisfaction

Very satisfiedSatisfiedNeutralDissatisfiedVery dissatisfiedTotal
Male30 (24%)50 (40%)35 (28%)8 (6%)2 (2%)125
Female25 (20%)60 (48%)30 (24%)8 (6%)2 (2%)125
Total5511065164250

Row % vs column %

Cross tabs support two percent modes you can toggle between.

  • Row percent — each row totals to 100%; shows the distribution within the row
    • e.g. satisfaction breakdown for males
  • Column percent — each column totals to 100%; shows the row breakdown within a column
    • e.g. male/female share among "Very satisfied"

Pick the mode that matches the question you are answering.

Supported combinations

Row / ColumnSAMAMatrixScale
SA
MA-
Scale-

Some combinations (e.g. MA × matrix) produce results that are hard to interpret and are not recommended.

Using URL parameters as axes

Values captured via the URL parameters feature can also be used as axes. Examples:

  • Satisfaction by traffic source (email / social / ads)
  • Purchase intent by campaign ID
  • External customer ID × behavioral data

Exporting results

Cross tab results can be exported as CSV or Excel. See Export for details.

Analysis tips

  • Focus on the size of the differences — meaningful gaps between segments are where insights live
  • Be careful with small cells — percentages from tiny n counts are unreliable
  • Combine multiple cross tabs to test hypotheses