The Feedback You’ve Rehearsed Four Times and Still Haven’t Given
It's Thursday, and you've had the conversation in your head again. Same person, same pattern, same imagined opening line. You've run it during your commute, in the shower, in the five minutes before your 1:1 with them starts. Then the 1:1 starts, you ask how their sprint is going, and the real conversation never happens.
This isn't a courage problem. You've had hard conversations before, with customers, with your board, with your own boss.
What's different here is that you don't have a structure you trust, so every version you draft in your head feels either too soft to matter or too sharp to survive. You end up saying nothing, which is its own kind of message: the behavior is fine, keep doing it.
Six weeks later it's still happening.
Now it's not just the original issue. It's the fact that you've let it run long enough that naming it feels like it requires an apology for the delay. The gap between "I noticed this" and "I said something" has become its own problem, stacked on top of the first one.
There's a decision buried in this too, even though it doesn't feel like one. Every day the conversation doesn't happen, the decision about what happens with this behavior doesn't disappear. It moves up. It becomes something you manage around instead of something they fix, which means you're now carrying a decision that should have stayed at their level. That's usually how it goes when feedback gets avoided: not that the issue disappears, but that the right to decide what happens next quietly escalates to whoever's still willing to carry it.
What changes when the structure exists
The managers who get through this fastest aren't the ones who found more nerve. They're the ones who stopped trying to improvise a difficult conversation in real time and started using a fixed shape for it: Situation, Behavior, Impact, Next Step — SBI+.
Situation anchors it to a specific moment, not a vague pattern ("in yesterday's client call," not "lately"). Behavior names what happened in observable terms, not character ("you cut off the client twice," not "you're not listening to clients"). Impact states what that produced: for the client, the team, the outcome. The "+" is the part most feedback skips entirely: a next step both people can point back to later.
Two to four sentences. That's it. The scripts that work are short on purpose. Length is exactly what a manager edits back in when they're nervous, padding the hard part with context nobody asked for. Specificity does the work that volume can't.
The same structure runs in both directions. A redirect script for the thing that needs to stop. A reinforce script for the thing that's working and should be named before it's taken for granted. Reinforcement avoided is still feedback avoided, just the easier-to-ignore kind. And a third script for the one that's been sitting unaddressed long enough that it's calcified into "just how they are": the pattern that's been worked around instead of named.
What's actually different afterward
The test isn't whether the conversation happens — managers have hard conversations all the time that change nothing. The test is whether the other person's response is different from the last three times you circled this same issue. Usually that difference shows up as specificity meeting specificity: they can tell you exactly what you meant, because you told them exactly what you saw.
The research backs the payoff. Dropbox's October study of its own workforce found that managers of its thriving employees were more than twice as likely to put AI time savings back into coaching, one-on-ones and development, and those employees were far more likely to say their performance feedback helped them improve (94% versus 75%). Betterworks' 2026 survey of 934 HR and business leaders named feedback as one of three central breakdowns in performance management, alongside measuring outcomes and goal alignment. The conversation you keep postponing sits right where most organizations already know they're weak.
The scripts get rehearsed once, with someone else in the room, before they're used live. Not because the words are complicated. They're not. But the first time you say a hard thing out loud is a bad time to also be improvising the delivery.
What does the conversation you've been rehearsing actually sound like, once you say it out loud to someone other than yourself?
If you want a clearer read on where this kind of delay is costing you, not just in this one conversation but across how decisions and feedback move through your team, the Executive Escalation Assessment takes under 12 minutes and gives you a specific picture of where the gaps are.
