Proxy Metrics

Proxy metrics are substitutes we use when we can't measure what we actually care about. Like counting hours worked instead of measuring productivity, or tracking website clicks instead of customer satisfaction. The danger comes when we forget these are imperfect substitutes and start treating them as the real goal. Think of it as mistaking the map for the territory - the measurement tool becomes more important than what it's supposed to represent. Reflect on this: Which metrics in your life might be proxies rather than the real thing you value?

Why We Use Proxies

We rely on proxy metrics because the things we truly value are often complex, subjective, or difficult to measure. Student learning quality is harder to quantify than test scores. Employee creativity is harder to track than hours logged. Customer loyalty is harder to measure than purchase frequency. Proxies give us something concrete to count and compare.

The Substitution Problem

The danger emerges when we substitute the proxy for the original goal entirely. In the school example, administrators likely wanted critical thinking, creativity, and lifelong learning skills - but measured these with standardized tests instead. When teachers focused solely on raising test scores, they may have reduced time for project-based learning, discussion, and curiosity-driven exploration that actually develop deeper understanding.

Common Traps

This pattern repeats everywhere: YouTube creators optimizing for watch time instead of educational value, call centers measuring call duration instead of problem resolution, police departments tracking arrest numbers instead of community safety. Each proxy creates perverse incentives that can undermine the original mission. The solution isn't eliminating metrics but aligning them more closely with what truly matters, using multiple complementary measures, and regularly questioning whether your metrics are serving your goals or becoming them.

Research in organizational psychology and economics has extensively documented Goodhart's Law and Campbell's Law as fundamental principles explaining why proxy metrics often fail. When a measure becomes a target, it ceases to be a good measure because people optimize for the metric rather than the underlying goal.

Key findings include:

• Goodhart (1975): When a measure becomes a target, it ceases to be a good measure [1]
• Campbell (1976): The more any quantitative social indicator is used for social decision-making, the more subject it will be to corruption pressures [2]
• Meyer & Gupta (1994): Performance measurement systems often create dysfunctional behaviors when metrics are misaligned with organizational goals [3]
• Bevan & Hood (2006): Target-setting in public services frequently leads to unintended consequences like gaming or neglect of unmeasured important aspects [4]

Proxy metrics aren't inherently bad - sometimes they're the best available option. The challenge is knowing their limitations. Some domains resist quantification entirely (art, relationships, wellbeing). In complex systems, multiple competing metrics may be necessary. Cultural context also affects which proxies work - what motivates people in one organization may backfire in another. Not all optimization is harmful; sometimes focusing on a proxy does drive genuine improvement, especially when starting from low performance levels.