Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.datagenie.ai/llms.txt

Use this file to discover all available pages before exploring further.

What the Correlation Matrix is

The Correlation Matrix maps the statistical relationship between every pair of KPIs in your dataset — including KPIs auto-connected across data sources by Nirvana. It surfaces which metrics are strongly correlated, weakly correlated, or inversely related — giving you the structural context to interpret stories more accurately.

How to read the matrix

Strong Positive

Values near +1.0 — the two KPIs tend to move in the same direction. A drop in one is likely accompanied by a drop in the other.

No Correlation

Values near 0 — the two KPIs move independently. A change in one carries no predictive signal about the other.

Inverse Correlation

Values near -1.0 — the two KPIs move in opposite directions. Useful for understanding trade-offs between metrics.

Practical uses

1

Identify co-movement pairs

Find KPIs that always move together. When one of these surfaces in a Top Story, check whether its correlated partner also shows movement.
2

Validate attribution

If KPI Attribution shows a side effect on a metric, use the Correlation Matrix to confirm whether this co-movement is structural or coincidental.
3

Build investigation hypotheses

Strong inverse correlations often reveal competitive or substitution dynamics between segments — e.g. as one channel grows, another declines.

Run the Correlation Matrix early when onboarding a new dataset. Understanding the KPI relationship structure helps you interpret stories more accurately from day one.

What to do next

Dimensional Analysis

Rank dimension values by Top N and Bottom N contribution for any KPI.

Data Playground

Build custom views and freeform comparisons.

KPI Attribution

Use correlation context to interpret contributors and side effects in a story.