Compare and rank

Radar / spider chart

Plots variables on radial axes and joins them to compare profiles.

  • Chart
  • Multivariate comparison
  • Profile comparison

What it answers

How do performance profiles compare across criteria?

Shape
Spokes and profile polygons
What it shows
Relative strengths and weaknesses.

Example

The question

How do our two products compare across six qualities customers rate?

Customer ratings out of 5 on six qualities, two products

Customer ratings out of 5 on six qualities, two productsSix spokes, scale 0 to 5 from the centre. Product A: Price 4.2; Quality 3.4; Delivery 4; Support 3.1; Design 3.6; Range 4.4. Product B: Price 3.1; Quality 4.5; Delivery 3.6; Support 4.2; Design 4.3; Range 3.Product AProduct BPriceQualityDeliverySupportDesignRange
Customer ratings out of 5 on six qualities, two productsSix spokes, scale 0 to 5 from the centre. Product A: Price 4.2; Quality 3.4; Delivery 4; Support 3.1; Design 3.6; Range 4.4. Product B: Price 3.1; Quality 4.5; Delivery 3.6; Support 4.2; Design 4.3; Range 3.Product AProduct BPriceQualityDeliverySupportDesignRange
Six spokes, scale 0 to 5 from the centre. Product A: Price 4.2; Quality 3.4; Delivery 4; Support 3.1; Design 3.6; Range 4.4. Product B: Price 3.1; Quality 4.5; Delivery 3.6; Support 4.2; Design 4.3; Range 3.

Illustrative data. It belongs to no company.

When to use it

Use it when

A few profiles with comparable bounded criteria.

Avoid it when

When polygon area or arbitrary axis order drives the conclusion.

How to read it

Read each spoke's scale and compare positions on the same spoke.

Note

Use comparable scales and an intentional axis order.

The data you need

Layout
Long table
Fields
Profile: text; Variable: text; Value: number
One row is
One variable value within one profile. What is the grain?
Rules
Define shared variable order and comparable ranges. Do not mix incompatible units without scaling.
Optional
Wide alternative: Profile,Speed,Quality.
Preparation
Align measurements by observation ID, select relevant variables and handle missingness. Scale or standardise only where the analytical method requires it.
Units
Each spoke needs a comparable scale or disclosed normalisation. Polygon area is not an automatic overall score.
Example of the input layout: Long tableA long table: one row per observation, few columns. New data adds rows.ProfileVariableValueAABBC
A long table: one row per observation, few columns. New data adds rows.

Worked example

A: Speed 80, Quality 90; B: Speed 70, Quality 95. With the same scale, A leads speed and B leads quality; do not compare polygon area as a score.

ProfileVariableValue
ASpeed80
AQuality90
BSpeed70
BQuality95