The Observer Effect in Performance Reviews
Our people analytics team has confirmed a long-suspected effect: the act of formally observing an employee's performance changes that performance. An employee who is unaware of being measured exists in a productive superposition of 'crushing it' and 'quietly browsing the glossary.' The moment a manager opens the review tool, that superposition collapses, and — critically — it does not always collapse the way the manager hoped.
This has uncomfortable implications for the annual review cycle. By measuring performance precisely once a year, we maximally disturb it precisely once a year, typically in the same week that compensation is decided. We now recommend continuous, low-energy observation instead: many gentle measurements that perturb the system less, in the same spirit as not staring directly at a wavefunction one wishes to keep intact.
We acknowledge the philosophical discomfort here. If performance only exists once observed, did the quarter happen? Our compliance office has ruled that, for payroll purposes, the quarter happened. For all other purposes, employees are encouraged to draw their own conclusions, ideally while unobserved.