Measure success without falling for ghost victories or gaming

Numbers can lie, even when they’re true. A city can declare victory on potholes because it filled more holes, while ignoring that complaints came mostly from rich neighborhoods. A school can celebrate higher graduation counts while learning has flatlined. These are ghost victories: impressive metrics that drift from the mission or reflect a rising tide.

Avoid them by pairing metrics. If you count quantity, pair a quality signal. Tickets closed must travel with repeat‑contact rates. Graduation counts must travel with independent assessments. When both look good, you’re closer to real success. When they diverge, you’ve learned something worth fixing.

Before you roll, pre‑game misuse. Ask your team, “If we hit these numbers but failed our purpose, what would we have done?” You’ll hear clever ways to game the system. Surface them now, then make them harder—tighten definitions, spot‑check, or add friction to bad shortcuts. Finally, track obvious external trends. If complaints fall everywhere because the weather improved or a platform changed, don’t overclaim. You can still learn, just credit the tide.

Behaviorally, paired metrics curb single‑minded target pursuit and Goodhart’s Law (“When a measure becomes a target, it ceases to be a good measure”). Pre‑mortems reduce the planning fallacy and surface hidden risks. Context checks protect you from false attribution. The result is a scoreboard you can trust—and that helps people improve rather than hide.

Pick one key result and add a balancing measure that speaks to quality or durability, then gather your team for a 15‑minute metric pre‑mortem asking how someone could hit the target while failing the mission. Use those answers to adjust definitions and checks, and add one context trend you’ll monitor so you don’t claim credit for a rising tide. Keep the set small and useful, then test it for a month.