Christine Marie

I have spent an embarrassing amount of my career writing feedback about feedback.

That’s what happens when you work in Human Resources. Eventually you find yourself on slide 37 of a presentation explaining to perfectly competent adults that feedback should be timely and specific, preferably beneath a diagram involving arrows.

I’ve taught the models, built the resources and coached leaders before difficult conversations. I have said things like, “What do you actually need this person to understand when they leave the room?” enough times that it should probably be embroidered on a pillow somewhere in my house.

I believe in feedback.

I am also excellent at avoiding giving it.

Not always. But I have absolutely stared at a message I needed to send someone and edited the feedback until it barely resembled feedback anymore.

You know this particular ritual.

You write what you actually mean.

This isn’t working. Here’s what I need you to change.

Then you read it and think, Ouch. That sounds harsh.

So you add something nice at the beginning. Now the transition into the actual feedback feels a little aggressive, so you soften that. Maybe “this isn’t working” becomes “I wonder if we might think about this a little differently.”

Much better.

Except now you’re worried they’ll think the whole thing is a big deal, so you add, “Nothing major!” Maybe a smiley face. Just to really destroy whatever credibility remained.

Twenty minutes later, you have produced an exceptionally warm, thoughtful message that leaves the recipient with absolutely no idea that you need them to do something differently.

And the worst part is, I know better. I teach this stuff for a living. Which is probably why something clicked for me recently while I was working with AI.

It gave me something I didn’t like, and I responded immediately.

No. That’s not what I mean. You’re focused on the wrong thing. Keep this section, lose that one, and stop using the same data point over and over to prove four different examples.

Which, for the record, is a real piece of feedback I have given AI while writing a proposal on training feedback. The irony is not lost on me.

It tried again. I told it what was better, pointed out what still wasn’t working and gave it an example. Then it tried again.

At no point did I spend twenty minutes wondering whether I had adequately acknowledged everything it had done well before telling it what it had done poorly. There was no carefully placed exclamation point to make the correction feel friendlier, and I certainly didn’t wait until our next one-on-one.

I just told it what it needed to know exactly when it needed to know it.

And that’s what bothered me.

Because I was doing, effortlessly, nearly everything I’ve spent years teaching people to do. I was specific about what wasn’t working and what I wanted instead, and then I gave it another chance.

No model. No carefully constructed conversation. No acronym. Why are there so many acronyms?

So why was it so easy?

The answer is obvious. An AI doesn’t have feelings. But I think that’s actually where this gets interesting, because the problem was never that we didn’t know how to give good feedback.

It’s that we were afraid to.

Not afraid of the feedback itself. We’re afraid of what happens in the three seconds after we give it.

You know the three seconds.

You finally say the thing you’ve been rehearsing in your head all morning, and then you watch the other person’s face change. Maybe they get quiet. Maybe there’s an “Okay” in a tone that has never, in the history of the English language, meant okay.

And suddenly every empathetic instinct in your body is screaming: FIX IT.

So you start talking again. A little more context. A reminder of everything they’re doing well. Whatever might bring the tension in the room back down.

I’ve done it.

I’ve watched someone’s face change after I’ve said something difficult and immediately felt the words lining up in my mouth to make it better. To make sure they understand that I still think they’re smart and capable and valuable and please don’t leave this conversation thinking I’m a terrible person.

And that’s the part worth paying attention to.

Because if I already know what I need to say, maybe the more useful question isn’t how do I give this feedback?

It’s:

What am I afraid will happen after I do?

I know my answers.

Maybe they’ll be upset with me or think I’m being unfair. Maybe I’ll have to stand behind what I said while they disagree with every word of it. Or, and this might be the one I like least, maybe they’ll leave the conversation liking me a little less than they did when they walked into it.

I don’t know about you, but I don’t love any of those possibilities.

And that’s the vulnerability in feedback we’ve built an entire training industry around, without spending nearly enough time acknowledging.

We talk a lot about the vulnerability required to receive feedback. But giving honest feedback requires some, too, because eventually I have to surrender control over what happens after I say the thing.

I can be thoughtful about how I say it. I can listen. I can make sure one piece of feedback doesn’t sound like an indictment of your entire existence. What I can’t do is control how you feel about it.

And if I’m being really honest, sometimes that’s exactly what all my softening, qualifying and reassuring is trying to do. I’m managing your reaction so I don’t have to sit in those three seconds.

But here’s the ugly truth: sometimes those three seconds don’t need fixing. Sometimes I just need to shut up and let another adult have a feeling.

That’s uncomfortable to write. It’s even more uncomfortable to do.

So what do we do about these icky three seconds we have suddenly found ourselves in?

I don’t know that I have an answer, but I do have…something. Somewhere along the way in my career – I genuinely can’t remember where I first heard it  –  someone told me, “Clear is kind.”

For years, I thought that was an instruction about communication: say what you mean and don’t make people guess where they stand. I’m not entirely sure that’s what it means to me anymore.

Clear is kind because clarity gives the other person something our discomfort often takes away from them: a choice.

If you tell me my work isn’t where it needs to be, now I know and can do something about it. If I’ve done something that affected you, give me the chance to repair it. Tell me what you need from me and let me decide whether I can give it.

I might not like the information. I might disagree with it. I might need to go home, complain about you to my husband, eat something with an unreasonable amount of cheese on it and come back tomorrow.

But at least I know.

The alternative is that you decided for me.

You decided I couldn’t handle hearing it. You decided now wasn’t the right time, that protecting my feelings (or let’s be honest, your feelings) mattered more than letting me know. And while you’re trying very hard to be kind, I keep walking around without information everyone else may already have.

Until eventually the consequences tell me.

That’s not kinder.

Now, let me pause here, dear reader, and admit something slightly inconvenient to my own argument: none of this is easy.

Feedback can be well–intentioned, come from someone I trust, and still feel like shit.

I work in Talent Development, so I would love to tell you that I receive feedback with grace and curiosity. That I lean forward, ask thoughtful questions and say something emotionally evolved like, “Thank you. You’ve given me a lot to think about.”

Sometimes I do.

Other times, I smile and nod while the tiny lawyer in my head starts assembling a defense, complete with exhibits and key witnesses.

Okay, but you don’t have all the context.

That’s not exactly what happened.

Sure, but did you consider…

I can get defensive when I feel misunderstood, then replay the conversation in the car and come up with several devastatingly articulate responses approximately forty–five minutes after they would have been useful.

And then, occasionally, something even worse happens.

Three days later, it clicks.

They might have been right.

Oof.

I hate those.

But some of the feedback that has changed me most is the feedback I didn’t receive beautifully. It’s the thing that irritated me enough to keep turning it over long after the conversation ended, until I could finally see the part I wasn’t ready to see when it was sitting right in front of me.

My reaction in the moment isn’t necessarily a measure of whether the feedback was worth giving. I don’t have to like it for it to be useful. I might disagree with half of it or wish you’d found a different way to say it. But none of that changes the fact that you’ve given me the choice what to do with it – and for that I am grateful.

Anyway, that’s what makes this whole thing so annoyingly human.

There is no perfect model that guarantees someone will hear exactly what you intended and react exactly how you hoped. Believe me, if there were, someone in Talent Development would have put it on slide 37 by now.

Which brings me back to that completely ordinary exchange with AI.

We know how to give good feedback. In fact, it turns out we’re pretty damn good at it when we’re not worried about what happens in those three seconds afterward.

So, thanks, AI, I guess.

You didn’t teach me how to give feedback. You just removed the part I was afraid of long enough to show me what I already knew.

The people in my life have feelings. So do I.

And that’s the part no feedback model can solve for us.

We have to care enough about another person to tell them what they deserve to know, and trust them enough to decide what they do with it.

They may not like it, but three days later, something might click. Or maybe it won’t.

Clear is kind.

It just isn’t always comfortable.