FactorPrism®
Use Case

The Region Hiding Bad News

Traditional analysis allocates impact proportionally. FactorPrism® finds the right level—separating what's company-wide from what's truly regional.

The Challenge

Understanding 12% Growth

A retail company's quarterly report showed revenue up 12% year-over-year. The executive team asked the obvious question: "Which regions drove this growth?"

This seems simple, but it's actually a hierarchy problem. Was it a rising tide lifting all boats? Or did specific regions outperform? The answer determines where to invest next quarter.

The Problem

Traditional Analysis Allocates at the Wrong Level

The finance team did what most companies do: allocate the 12% growth proportionally based on revenue share.

~
South (35% of revenue): +4.2% contribution Allocated proportionally by size
~
West (30% of revenue): +3.6% contribution Allocated proportionally by size
~
Northeast (35% of revenue): +4.2% contribution Allocated proportionally by size

The Hierarchy Problem

This approach allocates everything at the individual region level, assuming each region contributed proportionally to its size. But growth doesn't work that way. Some of it might be company-wide (affecting all regions equally), and some regions might be outperforming or underperforming that baseline.

By forcing all impact to the region level, traditional analysis can't see either the company-wide trend or the regional deviations from it.

The Reality

FactorPrism® Finds the Right Levels

FactorPrism® decomposed the 12% growth into its true hierarchy:

~
Segment-Wide Lift: +6.0% A rising tide affecting all regions equally (company-wide level)
+
South Regional Outperformance: +5.0% Growth above the baseline (region-specific level)
+
West Regional Outperformance: +4.0% Also beating the baseline (region-specific level)
−
Northeast Regional Underperformance: -3.0% Actually dragging vs. the baseline (region-specific level)

The Right Hierarchy

Half the growth (6%) came from a company-wide lift—perhaps market conditions, a successful brand campaign, or seasonal factors affecting everyone equally. This is a segment-wide effect, not a regional one.

The other half came from regional deviations: South and West outperforming that baseline, while Northeast was actually underperforming. Traditional analysis said Northeast contributed +4.2%. Reality: it was dragging by -3.0%, completely hidden by the segment-wide lift.

The Difference

Why Hierarchy Matters

Traditional (Wrong Level)

  • Allocates all 12% to individual regions
  • Northeast: +4.2%
  • Conclusion: "All regions contributing"
  • Action: Continue current strategy
  • Misses both the segment-wide trend AND the regional problem

FactorPrism® (Right Levels)

  • Separates segment-wide from regional
  • Northeast: -3.0% vs. baseline
  • Conclusion: "Northeast hiding behind company-wide lift"
  • Action: Investigate Northeast immediately
  • Finds the right level for each effect
The Outcome

Taking Action at the Right Level

With the true hierarchy revealed, leadership made targeted decisions:

High Priority
Investigate Northeast—found new competitor taking share, masked by overall market growth
Strategic
Document what's driving segment-wide lift to sustain it; document South/West success to replicate
Quick Win
Reallocate resources to defend high-performing regions while fixing Northeast
Problem Caught Early The regional VP identified competitive pressure in the Northeast and adjusted strategy before the segment-wide lift faded and exposed the problem.

Key Insight

Traditional analysis forces all impact to one level (proportional regional allocation). FactorPrism® finds the right level for each effect—separating what's truly company-wide from what's regional outperformance or underperformance. When a struggling region hides behind a rising tide, only proper hierarchy decomposition reveals the truth.

What's hiding at the wrong level in your data?

Connect your Snowflake data and find out in minutes.

Get it on Snowflake Marketplace