Annual reviews are the wrong unit of feedback

The annual performance review is one of the last rituals of industrial-era HR still standing. It survives partly out of habit and partly out of a belief that a formal moment is needed to make decisions about pay and promotion. The trouble is that the same moment is also being asked to deliver something it cannot: useful feedback that actually improves behaviour.

Behavioural science is unambiguous on this. Feedback is most effective when it is close in time to the behaviour it addresses, specific rather than global, and delivered in a relationship of trust. A comment about a presentation delivered eight months later has almost no chance of changing what happens in the next presentation. Worse, meta-analytic evidence shows that a third of feedback interventions actually reduce performance, particularly when they focus on the person rather than the task.

The alternative is not the absence of a review but a different cadence. Companies that have moved to continuous performance systems typically pair short weekly check-ins with quarterly written reflections and a lightweight annual summary used only for calibration. In controlled studies this pattern produces higher engagement, faster development and, notably, more accurate ratings because managers rely on recent evidence rather than reconstructed memory.

For HR the shift is more cultural than technical. The tooling exists. What has to change is the manager's default: replacing the once-a-year evaluation mindset with a weekly coaching one. That takes explicit training, protected time and a review of promotion and pay processes so they no longer depend on the annual document. Done well, the annual review becomes a summary of a year of good conversations, not the substitute for them.

Frequent feedback does not mean constant commentary. Too much unspecific advice creates noise and can make employees feel monitored. A useful conversation begins with the shared goal, identifies an observable behaviour and explains its effect. The employee should have space to respond and agree on a next step. Positive feedback deserves the same precision as corrective feedback; saying exactly what worked helps a person repeat it and shows that recognition is based on evidence rather than personality.

A continuous system still needs documentation and fairness. Important expectations, development commitments and performance concerns should be recorded in a consistent way. Quarterly reflections can help employees and managers review evidence before recency bias takes over. Calibration remains necessary for pay and promotion, but the discussion should draw on multiple examples gathered across the year. HR should audit outcomes for unexplained differences between groups and ensure that accessibility, leave or remote work do not reduce visibility unfairly.

Implementation can start without replacing every process. Define a short structure for one-to-ones, teach managers how to ask for the employee's view first, and protect time for the conversation. Then examine whether the existing annual form duplicates rather than supports those discussions. Track perceived usefulness, clarity and follow-through, not simply meeting frequency. A check-in that is repeatedly cancelled or contains no actionable exchange does not count as quality feedback. The goal is timely learning and fewer surprises, with the annual record reflecting conversations that already happened.

Employees also need the ability to request feedback and challenge an interpretation. A two-way process reduces dependence on a manager remembering the right moment and makes expectations easier to clarify. Teams can agree when feedback is best delivered, especially across cultures and communication preferences. For sensitive topics, privacy and preparation matter more than speed. Timely feedback means soon enough to learn, not so immediate that emotion replaces a fair account of what happened.