The feedback framework that actually changes behavior (and why most feedback doesn't)
Most feedback gets a nod, a thanks, and no actual change. The problem usually isn't what was said. It's that the behavior being described was never specific enough to act on.
Most managers have had this exact moment. You give someone feedback, they nod, they say something like "thanks, good point," and a week later you're watching the exact same behavior, wondering if any of it actually landed. The instinct is to conclude the person wasn't receptive. Usually, the real issue is upstream of that: the feedback itself was too vague to act on in the first place.
This isn't a small gap. Manager quality is one of the largest drivers of team engagement, and feedback conversations are consistently one of the most important levers within that. Getting feedback right isn't a soft skill in the abstract sense. It's one of the highest-leverage things a manager does.
Why vague feedback doesn't change anything
"Be more proactive." "Communicate better." "Work on your ownership." These phrases feel clear to the manager saying them, because the manager has a specific memory attached to them. The person receiving the feedback usually doesn't have access to that memory. What they hear is a vague directive with no clear next action, which means even a fully willing employee often has no concrete idea what to actually do differently.
The fix isn't softer language. It's more specific language. Feedback that describes a specific, observable behavior gives someone something they can actually change. Feedback that describes a general impression or a character trait doesn't, because there's no action attached to "be more proactive" the way there is to "loop me in before finalizing a client email."
People can change a behavior. They can't change being told who they are. "You're disorganized" is unwinnable. "The last two reports came in without the summary section" is something a person can actually fix.
A framework worth using: situation, behavior, impact
One of the more durable frameworks for structuring feedback breaks it into three parts, sometimes called SBI: the specific situation, the observable behavior, and the impact it had. The value of this structure is that it removes interpretation from the equation almost entirely.
Situation
Name the specific moment. "During yesterday's client call" is concrete. "You always do this" is not, and invites an argument about whether "always" is actually true.
Behavior
Describe only what a camera would have captured. "You cut off the client twice before they finished their question" is factual. "You seemed dismissive" is an interpretation, and interpretations are exactly what triggers defensiveness.
Impact
Explain what actually resulted. "The client had to repeat themselves and the call ran long" connects the behavior to a consequence the person can understand and care about.
Why specificity lowers defensiveness instead of raising it
It might seem like more detail would make feedback feel more pointed. The opposite tends to be true. A vague accusation like "you seemed defensive" invites an argument, because it's a claim about someone's internal state that they're perfectly positioned to dispute. A specific, factual observation like "you cut off the CFO twice during the Q&A" is much harder to argue with, because it's simply describing what happened.
Feedback that's grounded in a specific observation keeps the conversation anchored in reality instead of spiraling into a debate about intentions or character. That's a large part of why specificity, not softness, is what actually reduces defensiveness in these conversations.
Why timing matters as much as wording
Feedback that shows up weeks after the moment it describes has already lost most of its power. The person may not clearly remember the specifics, and the delay itself sends an unintended message: that the issue wasn't significant enough to raise when it actually happened. Feedback delivered close to the moment it describes is both more accurate and more credible.
This is also where a habit of writing things down as they happen pays off well beyond the moment of the note itself. A manager who jotted a quick, specific line right after noticing something, "cut off the client twice on today's call, seemed rushed", has exactly the raw material an SBI conversation needs, whether that conversation happens the same day or gets folded into a scheduled one-on-one a few days later.
The step most managers skip: closing the loop
A single feedback conversation, however well delivered, rarely produces lasting change on its own. Sustainable behavior change requires reinforcement, which means the conversation isn't really finished until a follow-up has been scheduled. Feedback treated as a one-time event tends to fade. Feedback that gets revisited, even briefly, is what actually sticks.
This is where a written record does double duty. It's not just documentation of what was said. It's the thing that makes the follow-up conversation possible at all. Without a note describing what was actually raised and when, a manager checking in three weeks later is relying on memory to describe a conversation that person might remember quite differently.
What ownership looks like versus compliance
The best feedback conversations don't end with the manager prescribing the fix. They end with the employee naming their own next step. Feedback that ends with the manager assigning a solution tends to produce compliance at best. Feedback that ends with the employee identifying their own change tends to produce something closer to real ownership, and ownership is what actually makes a behavior change stick past the conversation itself.
Before your next feedback conversation: Write down the specific situation and behavior you want to raise, in language specific enough that a stranger could picture exactly what happened. If you can't get that specific, the conversation probably isn't ready yet.