EP021
Normalize performance metrics for contextual factors before cross-unit comparison to ensure data-driven decisions reflect genuine process issues.
CyberTRIZ analysis · Process contradiction EP021 · one of 8,235 worked contradictions published by CyberTRIZ.AI
Regulations
Faster Identification of Underperforming Processes vs. Fair Comparative Analysis
Business Context. Quickly identifying which processes are underperforming allows organizations to focus improvement resources efficiently, but comparing processes across different contexts, such as different regions or customer segments, without adjusting for those contextual differences can produce unfair or misleading comparisons.
Process TRIZ Resolution. Rather than comparing raw performance figures directly, organizations should normalize performance metrics for known contextual factors before comparison, ensuring that identified underperformance reflects genuine process issues rather than uncontrolled contextual differences.
Applicable TRIZ Principles
Principle 3 (Local Quality) normalizes performance metrics for known contextual differences before comparison.
Principle 23 (Feedback) validates comparative findings against known contextual factors before acting on them.
Principle 35 (Parameter Changes) adjusts comparative benchmarks to account for measurable contextual variation.
Expected Outcome
Fast identification of genuine underperformance
Fair, contextually adjusted comparisons
Better-targeted improvement investment
Reduced disputes over comparative rankings
Decision Indicators
Process rankings do not account for known contextual differences between units.
Teams dispute comparative rankings as unfair given their operating context.
Improvement resources have been misallocated based on unadjusted comparisons.
No normalization occurs before processes are compared across contexts.
Underperformance labels have demotivated teams operating under harder conditions.
If several of these indicators are present, the contradiction is likely active and the Process TRIZ resolution above should be evaluated.