Most performance reviews don't fail because managers are incompetent. They fail because of patterns that repeat cycle after cycle, patterns that are easy to fall into and hard to spot from the inside.
Here are five of the most common ones, why they happen, and what to do instead.
1. Rating everyone in the middle
Walk into most performance review calibration meetings and you'll find a familiar problem: a huge cluster of "meets expectations" ratings and very few at the top or bottom. This is called central tendency bias, the instinct to avoid extremes.
It happens because managers want to be fair, and assigning a high rating to one person feels like it comes at someone else's expense. So they compress ratings toward the middle to sidestep that tension.
The result is a calibration that tells employees almost nothing. A "3 out of 5" at your company could mean anything from genuinely strong work to coasting.
The fix: Before ratings go in, show managers the distribution of their ratings compared to the rest of the organization. Peer pressure, visible, data-backed, changes behavior. Calibration sessions also help: when managers have to defend ratings to each other, inflated middles get questioned.
2. Evaluating the last 90 days instead of the full year
Recency bias is one of the most documented problems in performance management, and one of the least fixed. A project that went well in November gets weighted more than eight months of solid contribution. A rough patch in Q4 tanks an otherwise strong year.
The reason is simple: recent events are more vivid. Managers can't recall what happened in February with the same clarity as what happened last month.
| What recency bias looks like | What it misses |
|---|---|
| Strong Q4 → inflated rating | Q1–Q3 struggles go unaddressed |
| Rough Q4 → deflated rating | Strong full-year track record ignored |
| Last project outcome drives score | Long-term patterns invisible |
The fix: Continuous documentation. Managers who log notable contributions, outcomes, and feedback throughout the year, even brief notes, make far better reviewers than those writing from memory at the end of the cycle. Build this into your process, not as extra work, but as a default.
3. Feedback that's too vague to act on
Phrases like "needs to improve communication" or "should be more proactive" show up in review after review. They feel like feedback. They aren't.
Vague feedback fails employees twice: it doesn't tell them what specifically went wrong, and it doesn't give them a target to aim at. "Improve communication" could mean anything from writing clearer emails to speaking up more in meetings to managing up better. Without specifics, employees often guess wrong about what to change.
The fix: Anchor feedback to observable behavior and specific situations. Not "needs to communicate better," but "in three of our Q3 project check-ins, you sent status updates after the deadline, which left the team without information they needed to plan." That's actionable. The employee knows exactly what happened and what would be different.
4. Treating the review as a form, not a conversation
For many managers, the performance review is a document to complete. They write up the rating, submit it, share it with the employee in a 30-minute meeting, and move on. The employee leaves knowing their score but not much else.
This approach misses the point. The written review is supposed to start a conversation, not replace one. Employees want to understand how their manager sees their work, where they fit in the organization, and what their path forward looks like. A form doesn't answer those questions.
The fix: Make the conversation the primary deliverable. Send the written review in advance so employees can read and react before the meeting. Then use the meeting for dialogue, not delivery. Ask what the employee thinks, where they disagree, what they want to work on next. A review that ends with the employee having more clarity than when they started is a good review.
5. No follow-through after the cycle ends
This is where most review processes fall apart. Managers spend two weeks on reviews, calibration happens, ratings are shared, and then nothing. The development goals written in the review never get mentioned again until the next cycle.
Employees notice this. When goals don't get followed up on, the message is that performance reviews are administrative exercises, not genuine commitments. Trust in the process erodes fast.
The fix: Treat the review as the start of an ongoing conversation, not the end of one. Development goals should show up in 1:1s. Check-ins at 30 and 90 days after the review give managers a structured reason to revisit what was said. The cadence doesn't have to be heavy, a quick "how's the goal on X going?" in a regular 1:1 is enough to signal that the review actually mattered.
The pattern underneath all five mistakes
Look at these five mistakes together and a theme emerges: they're all about taking shortcuts when the process feels hard. Rating people in the middle avoids conflict. Evaluating the last 90 days avoids the effort of year-round documentation. Vague feedback avoids difficult conversations. Treating reviews as forms avoids the messiness of real dialogue. Skipping follow-through avoids accountability.
The fix isn't adding more process. It's making the right behavior easier than the shortcut. Better tooling, clearer expectations, and a calibration process that surfaces these patterns before they become norms, that's what moves the needle.
Confirm is built around exactly this: surfacing where reviews are going sideways and giving managers the structure to do better. If you're seeing these patterns in your organization, talk to us.
