Confidence shading

Confidence shading uses colour-coding to show how confident you can be in your profile results. It helps you distinguish between statistically reliable findings and those that might be due to chance or small sample sizes.

Note: Even with confidence shading turned off, the underlying statistical reliability doesn't change, you're simply choosing not to visualise it.

Note: Confidence shading is available on all profile types.

Interpreting confidence shading

High confidence - Red

These characteristics are very likely to be genuinely over or under-represented in your analysis group, not due to chance.

Prioritise these characteristics in targeting strategies, persona development, and campaign design.

Example: ‘Retired’ shows an index of 324 with dark red shading. You can be highly confident that retired individuals genuinely appear more frequently in your analysis group than in your base.

Moderate confidence - Orange

The pattern is likely real but has some uncertainty.

Consider these characteristics as supporting evidence alongside high-confidence findings. Use caution if making them primary targeting criteria.

Example: ‘Director’ shows an index of 143 with orange shading. The over-representation is likely real but less certain than higher-confidence findings.

Low confidence - Yellow

The pattern might be due to chance, small sample sizes, or insufficient data.

Don't base strategic decisions solely on yellow-shaded results. Investigate why confidence is low (often small sample sizes) and whether the finding makes logical sense.

Example: ‘Manual Worker’ shows an index of 567 with yellow shading. While the index is high, the small sample size (only two people in the category) makes this finding unreliable.

Common confidence patterns

Some common confidence patterns include:

• High index, high confidence

  This is the ideal scenario. It shows a characteristic is both distinctive and statistically reliable. These are your strongest findings.

• High index, low confidence

  The characteristic appears distinctive but lacks statistical reliability, often due to small sample sizes.

• Moderate index, high confidence

  The characteristic isn't dramatically over-represented but the pattern is statistically solid.

• Low index, high confidence

  The characteristic is confidently under-represented.