When the work isn't wrong, it's just not right

What it says

Kluger and DeNisi gathered 607 measured effects across 23,663 observations — the largest look anybody has taken at whether feedback works. On average, it helps. In more than a third of them, performance got worse. Their explanation is about where a person's attention ends up: feedback that keeps somebody looking at the task tends to help, and feedback that moves their attention onto themselves tends not to. The same words can do either. Which one you get isn't entirely up to you.

Where it comes from

A 1996 review in Psychological Bulletin, volume 119. It's the most cited thing in the field and it's unusually honest about its own limits — the authors say plainly that which features of a task decide the outcome is still poorly understood. That was thirty years ago. It hasn't been fully answered since.

What to do with it

Treat it as a conversation that can go either way, rather than a thing you deliver. Before you start, ask what this will have them thinking about an hour afterwards: the work, or themselves. If the answer is themselves, you're in the third that makes things worse, and more preparation won't save it. Different words might.

When it's the wrong lens

When the problem is that nobody has said anything at all. This one is about how feedback lands, which also makes it an excellent reason to keep putting the conversation off. Doing it imperfectly beats that.

What it says

Useful feedback answers three questions: where am I going, how am I going, and where to next. Most feedback answers the middle one and leaves the other two to be guessed at, which is why it so often lands as a verdict. They also separate four things feedback can be aimed at — the task, the way it was done, the person's own read on their work, and the person. Aimed at the person it's the weakest of the four. That goes for praise too.

Where it comes from

A 2007 review in Review of Educational Research. It's classroom research, written about students and teachers, and the jump to adults at work is one we're making rather than one the paper makes. Worth saying out loud, because the three questions are so usable it's easy to forget where they came from.

What to do with it

Write your three answers down before the conversation. If you can't answer the first one — where this is meant to end up — then it's early, and it'll turn into a conversation about the person, because that's the only subject left in it. Say the third one out loud even if it seems obvious. Obvious to you isn't obvious to somebody who has just been told their work has to change.

When it's the wrong lens

When what the work needs isn't a correction but a decision nobody has made yet. Then this turns an unmade decision into your coaching problem.

What it says

A shared belief that this is a group where you can say the risky thing — own up to a mistake, ask the obvious question, disagree with the most senior person there — without it being held against you. Edmondson studied 51 teams and found that safety predicted whether they learned, and the learning explained why the safe ones did better. The word people skip is shared. It's a property of the group, not a feeling one person has.

Where it comes from

Amy Edmondson, 1999, in one manufacturing company, measured with a seven-item survey the field still uses. It shows things moving together rather than one causing the other. It has also become a phrase people say in meetings, so whoever you're talking to may have heard it used badly. Don't lead with the term.

What to do with it

You can't announce it. What you can do is remember what happened the last three times somebody owned up to a mistake near you, because that's what everybody else is going by. If you want the conversation you're about to have to go well, most of the useful work has already happened — in how you took the last piece of bad news you were given.

When it's the wrong lens

As the explanation for one difficult conversation. A whole group's climate is a lot to reach for when one person needs telling that one piece of work has to change.

The move

Before the conversation, finish this in one sentence: "What good looks like here is ____." Then open with it, before you say anything about the version in front of you.

Why it works

It gives both of you somewhere to go. A gap between here and a named destination is a problem two people can work on. The same gap with nothing named is a complaint about somebody's work.

If you can't finish it

Then it's early. What you've got is a feeling that something's off, which is worth saying as exactly that — rather than dressed up as feedback it isn't ready to be yet.

The move

Point at something that exists — this paragraph, this screen, this number. "The third section assumes the reader already knows what we do", rather than "you write for insiders".

Why it works

A specific thing on a page can be changed this afternoon. A habit is a claim about who somebody is, and it gets defended like one. That's the move from the work to the person, and it's where a third of these conversations go backwards.

How it fails

Giving the example and then generalising from it anyway. "This section does X — and, honestly, you do that a lot" undoes the whole thing in half a sentence.

The move

Ask how it got here before you say where you want it to go. "Talk me through how this came together" — then let the whole answer happen, including the slow bit in the middle.

Why it works

Often enough, the reason is a constraint you didn't know about and the note that was ready to go was aimed at the wrong thing. Other times the answer names the problem itself, and now the two of you are agreeing instead of one of you correcting.

How it fails

Asking when you've already decided. People can tell, and a question you're not listening to the answer of is worse than not asking, because it makes the next one unbelievable too.