How to Give Constructive Feedback in Performance Reviews
Most managers dread the constructive feedback part of a performance review. Not because they don't have things to say—they do—but because they don't know how to say them in a way that lands well. So they soften until the message disappears, or they're direct to the point of being harsh, and neither actually helps the employee grow.
There's a better way. A simple framework, applied consistently, turns hard-to-say feedback into something clear, fair, and actionable. Here's how to use it.
Review season coming up? Kudo helps you write complete, specific performance reviews—including the constructive parts—in about 20 minutes per person.
Why Constructive Feedback Matters
Constructive feedback isn't a gift you reluctantly hand someone to justify their score. It's the most useful part of a performance review—for both sides. Without it, high performers plateau. Employees with blind spots never develop them. And managers who only deliver praise eventually lose credibility.
Done right, constructive feedback tells someone exactly what to change, why it matters, and what success looks like from here. Done wrong, it leaves them confused, defensive, or demoralized with nothing actionable to hold onto.
The difference is almost always structure.
The Framework: Situation → Behavior → Impact → Next Steps
This four-part structure is the most reliable tool for delivering feedback that actually lands. It's grounded in observable facts, not judgments, which makes it harder to argue with and easier to act on.
Situation
When and where did this happen? Ground the feedback in a specific moment or pattern, not a vague impression. "In Q3" or "during the launch kickoff meeting" is concrete. "Sometimes" is not.
Behavior
What did the person actually do or say? Stick to observable actions—not intentions, personality, or attitude. "Presented recommendations without supporting data" is a behavior. "Wasn't thorough" is a judgment.
Impact
What happened as a result? This is where the feedback earns its weight. Explaining impact—on the team, the project, the customer, or the business—turns an observation into a reason to care and change.
Next Steps
What does better look like? Close every piece of constructive feedback with a specific, forward-looking ask. This transforms feedback from a verdict into a direction.
The S→B→I→N structure keeps the feedback anchored in fact, not opinion. It's harder to get defensive about a behavior someone observed than about a character trait someone assigned. And the "Next Steps" element means the employee leaves with something to do, not just something to feel bad about.
6 Constructive Feedback Examples
Here's the framework applied to real situations managers run into every review cycle. Each example shows the full S→B→I→N structure in action—and a version of what the feedback might actually look like written out.
Example 1: Missing deadlines under pressure
During the product launch in September, Marcus missed two promised deliverables by 48 hours each. When timelines slipped, the team didn't learn about it until the day of the deadline. That compression forced two other team members to stay late and rework integration code at the last minute. Going forward, I want to hear about timeline risk the moment it surfaces—not after the deadline has passed. Flagging early is never a sign of weakness; it's what allows us to replicate across the team.
Example 2: Avoiding difficult conversations with peers
On at least three occasions this year—most clearly during the roadmap planning process—Priya declined to push back when she disagreed with a direction in the group setting, and instead raised her concerns with me privately afterward. The impact is that her perspective doesn't shape decisions in real time, and it signals to the team that disagreement isn't welcome in the room (even when it is). I want to see her voice concerns in the meeting, not after it. Disagreement is healthy here; avoidance isn't.
Example 3: Strong individual contributor not yet delegating
Throughout the year, Daniel consistently produced excellent work but absorbed work that should have been owned by junior team members. In Q2, for example, he rewrote two features rather than coaching the engineers who drafted them. His work was better—but the juniors didn't grow, and Daniel's bandwidth stayed artificially constrained. At Daniel's level, the measure of success is team output, not personal output. Next cycle, I want to see him deliberately hand off projects and coach rather than fix. Delegate at least one feature end-to-end, even when it's uncomfortable.
Example 4: Work quality is inconsistent
Sophie's output this year ranged widely—her best work (the onboarding redesign) was excellent, but her data integrity audit had significant gaps that required another team member to redo sections. The inconsistency makes it hard to calibrate what level of review is needed before her work ships. I want to see Sophie apply the same standard of care to her full workload, not just to the projects she finds interesting. A useful self-check: before handing anything off, ask "Is this the level of work I'd want to receive?"
Example 5: Technical skill gap affecting the team
This year, Chris struggled with database query optimization, which caused performance issues in two features that required emergency fixes after launch. When asked about his approach during post-mortems, it was clear this is a gap in his current toolkit, not a one-time mistake. This is an area to invest in directly. I'd like Chris to complete a structured course on query optimization in Q1 and pair with a senior engineer on at least two data-heavy features before owning one independently. The skill is learnable; the priority is addressing it proactively rather than discovering it through production incidents.
Example 6: Development goal for a high performer
Aaliya had an exceptional year by nearly every measure. One area to push into: executive presence. When she presents to the leadership team, she over-hedges—more qualifiers than the data warrants, which undercuts the confidence her analysis deserves. The insights are strong; the delivery needs to match them. For next cycle, I want Aaliya to lead at least two leadership presentations and practice leading with the recommendation before the reasoning. Her instincts are right; she needs to trust them out loud.
Common Mistakes Managers Make
Even managers who know the framework fall into these patterns under time pressure or discomfort. Recognize them before you write, and cut them from your reviews.
1. The feedback sandwich
Praise → criticism → praise. This structure is so well-known that employees have learned to ignore the middle part. If the constructive feedback is important, give it its own section. Don't bury it.
2. Vague growth areas that can't be acted on
"Should improve communication" or "could take more initiative" tells an employee nothing. They don't know what to start doing Monday. Feedback without a concrete next step is just commentary.
3. Saving it all for the annual review
If constructive feedback is showing up for the first time in a written review, you've waited too long. Annual reviews should confirm patterns the employee already knew about—not introduce surprises.
4. Judging intent instead of behavior
"She doesn't care about quality" is a judgment about character. "She shipped three features with known bugs rather than flagging timeline risk" is a behavior. One invites defensiveness. The other invites reflection.
5. Softening until the message disappears
When managers are uncomfortable, they hedge: "There's maybe a small area where you might consider…" By the time they're done, the feedback has dissolved. If something matters enough to say, say it clearly. The employee deserves that.
6. Writing feedback that's really about someone else's complaint
If you're writing constructive feedback based on secondhand reports but can't point to a specific, observed behavior, either gather more information first or don't write it in the formal review. Unverified feedback in a written document is unfair and creates liability.
Write better reviews in minutes — try Kudo free.
Kudo interviews you about each employee's projects, wins, and development areas, then drafts a complete, specific review in your voice. Constructive feedback included—no hedging, no generic phrases. Most managers finish each review in about 20 minutes.
Try Kudo Free →No credit card required. First review is free.
Putting It Together: A Quick Checklist
Before you finalize any constructive feedback in a review, run it through this check:
- Is the feedback tied to a specific situation, not a general impression?
- Am I describing a behavior, not an attitude or personality trait?
- Have I explained why this matters — the actual impact on the team, project, or business?
- Does the employee have a concrete next step, or just a problem statement?
- Would I be comfortable delivering this feedback in person? (If not, it probably needs more specificity or a different framing.)
- Is there any surprise here the employee didn't already know about from ongoing conversations?
If all six pass, the feedback is ready to write. If any don't, you have a gap to close before it goes on record.
Constructive feedback isn't about delivering bad news. It's about showing someone exactly where their next level of growth lives. Done well, it's the most valuable thing you'll put in a review.
Related: How to Write Employee Performance Reviews in 2026 — the full process framework from start to finish. And for language you can lift directly into reviews, see our Performance Review Phrases & Examples for Managers.