It’s Not Complicated. Instructional Coaches Should Give Clear Feedback (Opinion)

If you care about school leadership or culture and don’t know Michael Sonbert, you really should. He’s been a teacher and a novelist, cut his teeth as a coach working in the nation’s largest turnaround school network, and then founded Skyrocket Education in 2016 and Rebel Culture in 2022. Today, he works on leadership development with schools, Google, Northwell Health Systems, and many more. I find him a terrific source of straight talk on school leadership. That’s why, when he sent me a recent note on instructional coaching, I asked if he’d mind expanding it for you all. He agreed. Here’s what he had to say.

—Rick

Jim Knight, founding senior partner at the Instructional Coaching Group, author of multiple books on instructional coaching, and architect of a widely adopted approach to instructional coaching, recently penned the piece, “Should Coaches Give Feedback? It’s Complicated,” for ASCD, the education publisher, technical-assistance provider, and all-around K–12 juggernaut.

The essay, by a hugely influential coach for a hugely influential outlet, argues that “top down” coaching (as he refers to it) for teachers from coaches is often ineffective. Knight says, “The coach shouldn’t tell the teacher what data means, but ask questions and listen, trying to think with the teacher.” He continues, “Top-down feedback, I began to realize, was very helpful when there was a clear right and wrong way to do a task, such as when my dad taught me exactly how to skate backwards . . . [but not] when I tried it to discuss the complex environment of teachers’ classrooms.”

I respect Jim Knight’s substantial contributions to education, but here he’s wildly off. Knight’s assertion supposes a false binary, whereby coaching is either “top-down” or more of an exploratory conversation between teacher and coach.

You see, there’s a third option that Knight is missing.

Schools need to have an agreed-upon vision for instructional excellence. Once that vision is clear (by the way, teachers can absolutely contribute to that vision), coaches don’t need to play guessing games with teachers but can instead compare what’s happening in the teacher’s classroom against the exemplar and then tell (yes, tell) the teacher, with compassion and kindness, precisely what needs to get better and what the teacher should do to get there.

The coach should then model the skill the teacher needs to improve upon and have them practice that skill multiple times, giving feedback throughout, until they begin to build automaticity.

Knight’s approach assumes that getting a teacher to a place of being highly effective is like trying to answer a confusing, ambiguous riddle. But it’s not. I’ve worked in hundreds of schools in the past 15 years. The trends in classrooms across the U.S. are staggeringly similar, and what to do about them is surprisingly straightforward.

Now, that doesn’t mean execution is easy. Getting into great physical shape is straightforward: eat well, exercise, and burn more calories than you consume. But executing on it, for most people, is pretty difficult. Similarly, getting a school or individual educator to a place of being highly effective is straightforward: You start with systems, move to culture, and then go all in on instruction. But, as any educator can attest, executing this can often be challenging and fraught with obstacles.

What’s the point of using all the rubrics and frameworks that schools “use” if, when we enter a teacher’s classroom, we act as if we have no idea what success looks like? What’s the point of all the trainings and conversations about instruction if, when we observe teachers, we behave like great teaching is an unsolvable mystery?

A basketball coach, even at the professional level, wouldn’t ever ask the team to look at the score at halftime and have them analyze why they’re losing by 20 points. Instead, the coach would have meticulous notes on the places where the team can do better and then share those things with the team. Because the path to success is so clear, the coach can give feedback on how effectively the team is rebounding, playing defense, moving the ball, and—of course—shooting.

It’s the same thing in schools. More nuanced, yes, and with far more variables. Still, despite so many school leaders and teachers thinking their challenges are unique to them, they’re not.

The meetings that Knight describes, instead of radically building teacher skill, are just conversations. I know this because I’ve observed dozens of them. In these meetings, the teacher and the coach talk. And while some learning may occur—and they may even agree on some next steps—without actual skill-based coaching, very little changes. Which is why so many school leaders across the country are having the same conversations with teachers in May that they were having in September.

If we adopt Knight’s approach, we are wasting valuable time: the coach’s time, the teacher’s time, but most importantly, the students’ time. Maybe Knight is trying to solve a different problem than my team and I are trying to solve. But in the schools where we coach, things are extremely urgent. Students don’t have weeks or months for adults to figure things out. They need excellent teaching right away. In some cases, their lives literally depend on it. And in a profession dominated by terms like “equity” and “fairness,” isn’t it more equitable and far more fair to make change for students as quickly as possible?

There may be places for an approach like Knight’s. But they’re few and far between. When teachers are expert planners and have incredible classroom culture, collaboration about deep student engagement makes sense. When teachers can’t get students to sit down, are teaching without measurable objectives, and not assessing student outcomes, the approach just doesn’t make sense.

Knight’s approach, and others like it, at best assume that teachers have the ability and bandwidth to analyze their own classrooms and decide upon the next steps that would change student outcomes. I haven’t seen evidence that this is the case for the overwhelming majority of teachers. The approach, at worst, assumes that teachers are fragile, overly sensitive, and unable to receive straight feedback. Again, not random feedback grounded in what the coach thinks or is feeling in the moment but precise feedback aligned to the school’s instructional vision.

Moreover, these approaches also let coaches off the hook for being experts who can analyze classrooms, collect the most pertinent data, and model agreed-upon best practices for teachers and coach them to improve. Why do schools have coaches if they don’t coach but instead pass the buck to overworked, undersupported teachers to essentially coach themselves?

The rationale I’ve received from people who use Knight’s model and others like it is that they’re great for building relationships with teachers. I haven’t seen evidence that this is the case. So many of the teachers I’ve spoken to, on the subject of this kind of coaching, express frustration about long, meandering meetings and feeling like they’re trying to guess the answers to the coach’s questions when the coach could simply tell them instead.

Imagine you were lucky enough to receive tennis lessons from Serena Williams. It would be exhausting to spend a good chunk of time playing Q&A about proper form, foot positioning, and ball placement. But it would be invigorating to be taught by an expert. It’d be thrilling to know that in a very short time, you’d be better because of her coaching.

This is where great relationships come from. When a coach’s feedback is spot-on and a teacher knows that implementing that feedback will improve their teaching, students (and their teachers) will be better off. Strong relationships and trust come from providing value for someone (quickly), not spending a teacher’s entire prep pretending that questions about good teaching are unanswerable.

To the question about whether or not coaches should give feedback, it’s not complicated like Knight asserts. They should. To do anything less is borderline negligent.