Questioning Coaching
Instructional Coaching is the current Big Thing in teacher development. But is it all it's cracked up to be, and what are the common barriers to its successful implementation?
Instructional Coaching seems to be everywhere at the moment. Books, blogs, working papers, tweets, conferences, podcasts and LinkedIn posts abound. As a model for teacher development, it is certainly the current Big Thing.
Enthusiasm for Instructional Coaching is, in my opinion, well-grounded and completely reasonable. I remember reading Dr Sam Sims’ blog on the topic back in 2018, which persuasively argued that Instructional Coaching is currently the “best evidenced” form of CPD. Sims and others have by now extensively charted that evidence base, and it is very clear that Instructional Coaching is not an unevidenced “fad” like learning styles or brain gym.
Despite this, there are a number of reasons to be hesitant. Whilst Instructional Coaching is well-evidenced, that doesn’t necessarily mean it will work in the messy reality of schools, and below are a number of questions and challenges that teachers and leaders can use to judge whether it is right for their context, and indeed whether it’s all it’s cracked up to be.
Challenge 1: It is psychologically fraught
On a visit to Schoolington Academy, the leaders tell me they have recently pivoted to Instructional Coaching for teacher development. They have split the staff body in two. Half the teachers were sent off to train to be coaches. Each coach will then be paired up with somebody else in the school who will be their coachee.
I ask them how that was received by staff, and they told me that it went down surprisingly badly, and that a lot of the coachees were quite upset. In addition, those upset weren’t the usual moaners, but were generally colleagues who were well motivated and wanted to get better at their jobs.
I’ve experienced variations of the above a number of times. Leaders stand up in the hall and say things like “everyone is going to be coached, and it’s about everyone improving regardless of their starting points,” and are subsequently surprised to learn that what colleagues hear is “you are not good at teaching, and need coaching.”
This occurs because according to most definitions of Instructional Coaching, you need to have one coach and one coachee, with differing levels of relative experience. By design, you are creating an imbalance in the pairing. One is the “expert” and one is the “novice.” As a leader, you might work to try and ensure that doesn’t lead to any ill will or hard feeling, but there is no doubt that you have set the field up with a tilt. Trying to convince people that it doesn’t imply anything about starting points or expertise is fighting against the very definition of the term “coach.” You have created a two-tier system, and should not be surprised that people placed on the second tier do not respond well.
By instituting Instructional coaching, you create a two-tier system, and should not be surprised that people placed on the second tier do not respond well.
This is even worse in schools where small groups of staff (who aren’t trainees) have Instructional Coaching, but nobody else does. This creates a three-tier system, where you have a first tier made of coaches, a second tier made of teachers who neither receive coaching nor have been selected to coach, and a third tier of teachers who need coaching.
This stratified macro social context is therefore primed for conflict, leading to micro social and psychological problems arising from fraught coaching conversations. Delivering feedback in an unbalanced system is not a straightforward endeavour, and the risk of upset, demotivation and offense is high.
In schools, these risks are heavily exacerbated by a number of concrete and often underappreciated factors like historical trauma around formal observation or the deep sense of insecurity that can come with being observed (rendered more acute if you are observed with difficult classes). We might consider it an unfair characterisation, but it should come as no surprise to us that some consider Instructional Coaching to be “performance management with better branding.”
These problems are not insurmountable. Given excellent training and school culture, it’s possible to assign some people as “coaches” and others as “coachees” and reduce the cultural impact of creating a two- or three-tier system. But it isn’t easy and it isn’t likely.
Challenge 2: Nobody knows what it is
The term “instructional coaching” struggles with a lack of a clear definition. Evidence Based Education’s Professor Rob Coe notes a number of conflicting definitions of the term, and even where some definitions are similar, the emphasis is often on different aspects and processes. In recent years, there has developed an ecosystem of IC-adjacent terms like “responsive coaching”, “cognitive coaching,” “coaching for adaptive expertise,” accompanied by multiple frameworks like GROW, Co-Active and MSTR (further discussion around terminology and approaches can be found in Josh Goodrich’s excellent book or Adam Robbins’ blog).
I don’t want to retread well-worn ground, but it is worth explaining why this is a problem. It’s not that any of these definitions are false or inaccurate or bad: nobody owns the term “instructional coaching,” and many of the ideas within this ecosystem are interesting and useful. The problem is that it leads to incoherence and bad implementation.
My personal experience of this is a case in point. I currently mentor two teachers in their early career training. Whilst I train them in my school, their training is accredited and overseen by external organisations. In my case, my two mentees are under different training providers, both of whom explicitly want me to use “instructional coaching”. However, whilst they both claim the term “instructional coaching”, the substance of what I am asked to do is radically different.
To cite just one example, one training provider asks me to focus my weekly observations on a pre-organised curriculum of specific areas (e.g. questioning one week, subject knowledge another) and then construct action steps myself. The other asks me to choose the focus myself, then pick an action step from a pre-written bank.
When it comes to observation focuses and action steps, these two providers are therefore asking me to do the exact opposite of each other. To then call both of these approaches “Instructional Coaching” renders the term almost completely meaningless.
I am being asked to do Instructional Coaching by two different providers, but they have radically different interpretations of the term.
When you scale this problem across the teaching population, we end up with significant implementation problems. School leaders and training providers are all using the same research papers, blogs and general thought world to build their programmes, but each is emphasising different (and often contradictory) findings from that thought world. This ground becomes ripe for “lethal mutation”, where an idea that appears to have a strong evidence base fails to be implemented correctly, potentially causing more harm than good. In the same way that school leaders took research on “feedback” and turned it into “written marking”, they will take research on “instructional coaching” and use it for all sorts of professional development purposes, which may or may not be effective.
My general opinion with such things is that “if a reasonably well-read teacher or school leader cannot get to grips with the precise meaning of a term, then the term is probably problematic.” Individual providers and thinkers might argue that they have good and consistent definitions of Instructional Coaching, and that’s probably true, but if we need to have done a huge amount of intellectual legwork to understand exactly what is being argued or proposed and how it is different from what the next person has argued or proposed, then lethal mutation and poor implementation is inevitable.
Challenge 3: It is expensive
It is not controversial to say that schools should spend money on teacher development. But they do not have unlimited budgets, and Instructional Coaching is a relatively expensive form of teacher development due to the time demands. There are a number of reasons for this, but I think the most underemphasised one relates to the time commitment: if I am to coach a colleague, I need at least an hour a week to do so, as I need to both observe and give feedback. That’s an hour that I’m not teaching, and the school needs to pay somebody else to do that teaching instead of me.
Let’s put some ballpark numbers to it. A typical secondary school might have ~60 teachers. If 20 are designated as coaches at one hour per week, that’s an extra 20 hours a week of lesson time that needs filling. There are 40 coachees, each of whom will require an extra half an hour to be coached in, totalling 20 additional hours per week.
That’s a total of 40 teaching hours a week that will need to get made up somehow, and with typical teachers teaching for 24 hours a week, we end up with an average secondary needing to employ 1.7 new teachers. With the median teacher salary being £51,000, that totals £86,700 per annum, without taking into account national insurance and other costs, and assumes that we can perfectly fit the timetable around subject demands. These calculations are rough, and should only be treated as such, but it gives us a picture of broad and approximate costs.
Of course, schools can reduce this cost by having fewer people coaching and being coached, though this exacerbates the two- and three-tier problems mentioned above. Schools could also have observations once a fortnight, or reduce the time required for those observations and feedback. Whether such reductions would result in the Instructional Coaching being effective is another question, though its logical to assume that less frequent coaching results in weaker impact (it’s worth noting that Bambrick-Santoyo’s work on coaching, which has influenced a lot of the contemporary Instructional Coaching discourse, advocates coaches observing coachees on a near-daily basis).
Schools could also reduce workload in other areas so that coachees can have feedback during PPA or other non-teaching time. This is easier said than done, and unfortunately in some cases that I’ve heard of this hasn’t happened, and it’s become just another thing teachers are expected to do with no extra time.
As with before, the cost is not an insurmountable hurdle. Schools could decide it’s worth it. They also could find ways to bring it down, but we are starting from a pretty high bar, and we should be honest and transparent about those costs.
Challenge 4: It relies on people being radiologists
Instructional Coaching simply cannot work if the coach is not a highly skilled observer of teaching. I’ve previously referred to such a person as a “radiologist”, someone who can perceive details and elements that would pass completely unseen by a more casual observer. I’m not aware of any hard evidence on how many coaches would meet the definition of “radiologist”, but my experience is that a great many would not.
There’s no blame or judgement attached to that claim. To me, the failure rests on our often inadequate teaching and leading training infrastructure, not the individual coach. But if the claim is true, or even close to true, then Instructional Coaching as a professional development tool enters the process at the wrong stage. There’s no point discussing what coaching and feedback sessions should look like if the coach cannot adequately understand what’s happening in the lesson.
Challenge 5: Most schools and teachers lack a shared understanding of what good looks like
I doubt there’s a teacher in the country who has not received feedback they disagreed with. In fact it’s so common, it became a meme, with teachers posting the most outlandish feedback they received in formal observations:
As a profession, we do not yet have a shared understanding of what “good” looks like. We might discuss and debate it, but we often don’t agree. In some cases, the differences are marginal. In others, they are significant. In many, they appear trivial but are actually substantive. Some differences relate to philosophies, and others might relate to different subjects and disciplines.
All too often, differences stem from the fact that we simply haven’t thought hard enough about the strategies and skills that make up good teaching. We read a blog or saw a video with a cool idea or teaching strategy, but we don’t understand how it has specific boundary conditions that dictate when it should or shouldn’t be used.
Because of this, despite having centralised Teaching and Learning policies, many schools are pedagogically anarchic, made up of a kaleidoscopic and incoherent milieu of philosophies, strategies and classroom microclimates, rendering meaningful engagement on improvement almost impossible.
For Instructional Coaching to work, you need your coaches to be expert observers, and to have a strong shared understanding of what good teaching looks like. In many contexts, this simply isn’t the case.
Some schools have opted for centralised action step banks in order to counter the range of pedagogical skill and ideology within their organisations. Theoretically, this can work, but it feels somewhat like putting the cart before the horse - are people going to adequately diagnose and prescribe action steps if they don’t really understand them?
To make our recurring point again, this problem is not insurmountable. But it’s extraordinarily difficult to fix, and school-wide IC seems doomed from the start if there is no shared understanding of “what good looks like.”
Challenge 6: Which teachers become coaches?
When choosing coaches, schools tend to look to their best teachers. This sounds perfectly reasonable, as we’d expect them to be radiologists and to have a strong understanding of what good teaching looks like. We do, however, have some problems with this logic.
Firstly, assuming we can identify the “best” teachers, we can’t guarantee that they would necessarily make the best coaches. A lot of great teachers do what they do intuitively and tacitly, and cannot necessarily recognise it in themselves, let alone in others. We could (and should) train them to be coaches, but we mustn’t underestimate how difficult that is, and how much time and effort might be required to make someone an effective coach. A twilight training session once a term or a single conference day does not a coach make.
Secondly, we also have to recognise that taking our best teachers out of the classroom isn’t value neutral, because it means that students will be taught by someone worse instead. This is a perennial deliberation, and one that vexes the education system at all levels (e.g. every time we offer a brilliant teacher a promotion to leadership we are moving them further away from the classroom). Making a great teacher a coach takes away a certainty (these children are getting a great education) and replaces it with an uncertainty (this coach can make other teachers great at teaching). On balance, we might decide it’s worth it, but as we’ve argued previously, we can’t pretend like there are no downsides involved.
Challenge 7: It’s not for everyone
Many of the IC programmes that I have seen seem designed for inexpert teachers: those who are new to the profession, or who are experienced but have had poor training, or who are experienced but used to very different contexts. In such cases, the gap in relative expertise between coach and coachee is large, which helps with some of the psychological and implementation issues above. More importantly, there is not just a relative difference, but an absolute one that leads us to favouring an IC model. For example, if a teacher is not using Name at End questioning, their coach can:
1. Readily observe this lack,
2. Communicate its importance,
3. Formulate an action step,
4. Have the coachee do deliberate practice
5. Observe it in the next lesson (or very soon).
Every single step here changes drastically if the absolute expertise of the coachee is higher. For example, in a recent video I posted I am building a longer answer, taking responses from students point by point:
The first student says “the particles move faster”
I ask another student for the next point in the answer
She says “they have more kinetic energy.”
I explain to her that that’s the same marking point
She gives me the next point (“there are more frequent collisions”) as a follow up.
Shortened section of the video above. Click here to see the whole sequence and discussion.
In hindsight, I should have asked her why her answer wasn’t a good addition to what the previous student said. This would have both ascertained if she was listening to the first student and whether she could recall that these are the same marking point.
Tracking back to our list above:
I received hundreds of comments on that video across various platforms, and to the best of my knowledge nobody observed this problem
If they had, they could have communicated it to me extremely quickly and easily, because it relates to an area of teaching I already know about.
I don’t need a formal action step or to go through the rigamarole of constructing one.
I certainly don’t need to go into an empty classroom to deliberately practise this, or spend time writing out example questions that relate to the same area of teaching.
It’s unlikely that a similar scenario will arise in the next lesson or any point soon.
Broadly, the standard coaching models would be the wrong approach for the feedback I needed. It could be that if I moved schools or made some other drastic changes in context this type of coaching would be good for me, but right now it isn’t. I know I can do better, and I know I have lots of things I need to improve on, but in this case Instructional Coaching won’t help me, it will slow me down.
Instructional Coaching adherents would argue that at this point you have to change the model, and they would never advocate for the situation I am describing. This is of course totally reasonable, but if a school is rolling out Instructional Coaching across the staff body (which is often what is actually happening), then this kind of inflexibility is either inevitable or going to lead to multi-tier systems, where some people follow the Instructional Coaching model and others don’t.
Again, are the problems insurmountable? No. Are they worth considering? Absolutely yes.
Challenge 8: It often makes the proxy the measure
When Instructional Coaching is rolled out, how do we know that it’s working? Often, schools and training providers focus on whether it’s happening: are teachers going into lessons? Are they coaching each other? Is the feedback being recorded? Broadly, are they engaging with the process?
To cite Rob Coe again, this would be a classic case of “mistaking school improvement.” The thing we are interested in isn’t “is Instructional Coaching happening?” That’s just a proxy. The thing we are really interested in is “is teaching improving?” Often though, the proxy becomes the measure.
One could argue that all professional development initiatives suffer from the same problem, and I’d agree. But I worry that Instructional Coaching, given its tendency towards intensive bureaucratic and formulaic elements, lends itself particularly to this kind of proxy-becoming-the measure. For example, in all the years that I have been implementing IC on behalf of a range of different providers, whilst I have (admittedly) had to be chased a number of times for technical and prosaic things that I haven’t done (like submit the right piece of paper to the right drive or give enough information in a particular box online), I have never once received feedback on the pedagogical quality of the advice I have given. I’ve only been observed actually coaching once, and when my coachees are observed by someone from the provider, the feedback is always about the coachee and what they can do better: nobody ever turns round to me and says “ok, so I saw x in the lesson, what’s going on with your coaching such that x took place?”
At risk of sounding like a broken record: this problem is not insurmountable. But it is the reality on the ground for many schools that have rolled out Instructional Coaching.
Is Instructional Coaching all it’s cracked up to be?
I do want to state again that I suspect none of what I’ve written above is unknown to adherents and promoters of Instructional Coaching. As I hope is clear by now, I strongly believe that Instructional Coaching is well-evidenced and can be an effective form of professional development. But it isn’t always, and that’s because there are a whole host of barriers standing in the way of its effective implementation.
In light of all the above, I think some of the breathlessness and wholesale adoption of Instructional Coaching is problematic. It’s a wonderful thing if you can get it to work, but that “if” is doing a lot of heavy lifting. As a leader, you might have tried to implement Instructional Coaching, and you might have found it didn’t go as well as intended. Given the eight problems presented above, I’m not sure how surprising that is, and I doubt you are alone. It’s now up to you to decide if it’s the right approach for you and your school, or if it’s time to try something else.
Disclaimer: many readers may know that alongside my teaching and school support work I co-own Carousel, and our Carousel Teaching CPD platform provides a direct alternative to Instructional Coaching. I therefore appreciate and respect the fact that not everyone will consider me a neutral observer. That said, I’ve tried hard in this blog to consider Instructional Coaching on its merits, and to be as objective as possible in my analysis.
Great post, Adam.
I'd say, in today's climate, challenge 3 is the biggest issue, especially hard to justify given the other challenges involved.
Of course if you were a bullish senior leader able to show ( using the proxies mentioned) some short term 'impact' that might help you on your way up the career ladder, it might not stop you. But that never happens.
Your scope creeps at points in this post: from “Some schools implement coaching poorly” to “Instructional coaching as a class of diverse models is problematic.”
That said, I think you’re directionally correct that IC tends to expose existing dysfunctions (psychological fragility, lack of shared instructional model, high-stakes observation norms, weak curriculum)… and, in the worst cases, can amplify those weaknesses. My read, though, is that this reflects naïve or incoherent implementation rather than a decisive critique of coaching per se.
With that reframing, I think this could serve as a warning document for leaders against careless rollouts. I plan to run it by my team as a diagnostic checklist of predictable failure modes we need to guard against (or address!) as we tighten our own model.