This isn’t about which appraisal method to use or how often reviews should happen, we’ve covered that ground in our guides on appraisal methods and why the annual-only model backfires. This is narrower and more practical: once you’re actually sitting across from an employee, what makes that specific conversation land well or badly.
A review built from memory a few days before the meeting is dominated by whatever happened most recently, a well-documented pattern of specific incidents (Critical Incident Reports, or even a simple running log) protects against this recency bias and gives you concrete examples instead of a general impression.
An employee should already know, from the beginning of the period, exactly what “good” looks like for their role. A review that introduces new expectations for the first time during the meeting isn’t a fair evaluation, it’s a surprise dressed up as feedback.
Leading with praise specifically to soften a criticism, then closing with more praise to end on a high note, tends to read as manipulative rather than kind once someone’s experienced it a few times, and it can bury the actual message so effectively that the critical part doesn’t land at all. State both the strengths and the specific areas for improvement clearly and directly, without using one to cushion the other.
“You need to communicate better” gives someone nothing to act on. “The client update for the Q3 project was two days late without a heads-up, and it created a scramble on our end” names a specific, verifiable instance and its actual impact. Specificity is what turns feedback into something someone can actually change.
A review that’s entirely manager-to-employee misses the employee’s own perspective on obstacles, priorities, and context the manager may not have visibility into. Build in real space for the employee to respond, not just acknowledge, before moving to the next point.
Negative feedback, even delivered well, can genuinely trigger defensiveness or frustration. That’s a normal reaction, not evidence you did something wrong, give it a moment rather than rushing past it, then steer the conversation back toward the specific behavior and the path forward.
A review that ends with “let’s keep working on this” gives nothing concrete to check against next time. A specific action, a deadline, and how success will be measured turns the conversation into something both people can actually follow up on.
If a raise or promotion decision is riding on this same conversation, both sides will unconsciously manage toward that number rather than toward an honest account of performance. We’ve covered why this specific bundling causes most of the negativity associated with performance reviews, and how to structurally fix it, in our guide on why annual reviews backfire.
“Let me know if you need anything” is easy to say and easy to ignore. Naming a specific resource, a training budget, a mentor, adjusted scope while priorities shift, signals actual investment in the outcome rather than a polite closing line.
Q: Should positive feedback always come before negative feedback in a review?
A: Not as a fixed rule. Ordering feedback specifically to soften criticism (the “compliment sandwich”) tends to dilute the actual message and can start to feel manipulative over repeated reviews. Present both clearly and specifically instead of using one to cushion the other.
Q: How much documentation is actually needed before a review?
A: Enough to support every point raised with a specific instance, not a general impression. A running log built throughout the period, even informally, beats trying to reconstruct months of performance from memory right before the meeting.
Q: What if an employee becomes upset or defensive during the review?
A: Let the reaction happen briefly rather than rushing to smooth it over or retreating from the feedback. Once it settles, steer the conversation back to the specific behavior and a constructive path forward rather than avoiding the topic going forward.
Q: Should self-assessments be part of every review?
A: They’re genuinely useful, having an employee document their own view of the period first surfaces their perspective before the manager’s assessment anchors the conversation, and it reduces the manager’s prep burden since some of the material is already assembled.
Q: How long should a performance review meeting actually take?
A: Long enough for a genuine two-way conversation with specific examples on both sides, rushing it to fit a fixed 15-minute slot usually means skipping the parts that actually change behavior. The right length depends on how much ground needs covering, not a fixed calendar default.
Q: Is it a problem if compensation decisions happen in the same meeting as developmental feedback?
A: It’s a common structural problem, not a personal failing of the manager running it. Bundling the two makes both worse, an employee anxious about pay can’t fully absorb developmental feedback, and a manager reluctant to jeopardize a raise decision tends to soften honest feedback that would have actually helped.
A performance review succeeds or fails less on the method behind it and more on the specific conversation in the room: documented, specific, genuinely two-way, and closed out with something concrete to act on. Get those elements right and the underlying method matters considerably less than most guides suggest.