March 27, 2026
Bottlenecks and maximizing opportunity to learn
Friends, thank you for subscribing to The Coaching Letter—you are awesome. A couple of recent Coaching Letters have been about bottlenecks and Goldratt’s Theory of Constraints (ToC), so this CL is about that concept applied to education—and specifically Opportunity to Learn (OTL).
To recap a couple of things: OTL is the very basic idea that if we want students to master a given set of competencies, they have to be provided with access to the salient inputs for them to be able to learn (CL #223). In our framing, there are three major inputs: time, task, and teaching, and these three inputs can be decomposed into their component parts—a high quality task, for example, comprises grade level content, knowledge rich content, appropriately challenging cognitive demand (i.e. requires students to figure something out), and is accessible to all students (low floor), but never runs out of challenge as students move through the task (high ceiling). Here’s an early version of how we started to visualize this.
So if we are right about how to think about OTL, then the longest lever we have to pull in both increasing achievement of all students and reducing variation in both experience and outcomes between our highest and lowest achieving students is to engage ALL students with challenging grade-level tasks, ensure the cognitive demand of the task is supported by effective instruction, and to give students adequate time, measured in terms of exposure to content.
For more on these three: I wrote a long Coaching Letter a few months ago about Task (CL #213—which was featured in the Marshall Memo), with the most hyperbolic tag line I could come up with: On the acute and critical need for urgent attention to Task. I wrote about how I learned about the role of cognitive demand from research that Tom put together in Core of the Matter #9. Since then, we’ve gotten better at understanding how cognitive demand functions within the instructional system—not just as a property of the task, but as something shaped by curriculum, beliefs, instructional practices, and leadership. Tom wrote about instructional time in Core of the Matter #10. I wrote about curriculum in CL #213. And I wrote about holding environment in CL #178 and CL # 179 (about Ted Lasso’s holding environment which, coincidentally, was also featured in the Marshall Memo).
We’ve been thinking about OTL through the lens of systems optimization and bottlenecks, thanks to a deeply nerdy (and unexpectedly helpful) book from 1984, The Goal, by Eliyahu Goldratt. The book was a very valuable tip from my BFF Kerry. It’s a novel, which is a bit strange (although not the only fictional account of organizational improvement—you’ve probably heard of Our Iceberg is Melting, for example, or Who Moved My Cheese). I am not claiming that The Goal is great literature—but it is oddly compelling. The book is about clarifying your goal (and, as Kerry put it, the goal of the goal) and it is also about identifying constraints in your system. But you keep reading not because you care deeply about throughput accounting, but because you want to know if Alex, the protagonist, can save his factory, his marriage, and his job.
I think we’ve all heard the maxim that every system is perfectly designed to get the results it gets. The point is that there’s no point in pointing at any one thing or person and blaming them for results. There is a system at work, composed of stocks, nodes, and flows, which interact to create outcomes. Education is a system. And I think we still get a lot wrong. We still explain outcomes in ways that locate the problem in students rather than their opportunity to learn and the conditions they experience. We talk about motivation as if it’s something that people bring with them to any given situation, rather than a belief that is developed as a result of experience of success or failure, belonging, and expectation. We think that kids who score poorly on standardized tests are less intellectually capable than kids who score well, even though we know that those scores are a product of multiple factors, and that despite the scores, they are all capable of doing challenging thinking.
And then we act on those beliefs. We assign easier work to the very students who have had the least opportunity to learn. We lower cognitive demand in the name of support. We fragment content, over-scaffold, and narrow the task so that students can complete it without having to think very hard. In doing so, we create the outcomes we claim to be responding to. The system interprets low performance as inability, and then produces instructional experiences that guarantee it will persist.
Yes, I know. Harsh.
If we were to think in terms of systems to improve student learning, we would optimize for OTL. In our work, we are getting better at helping schools and districts optimize for OTL, especially through the development of instructional practices and leadership routines. Still a long way to go, though… And what has been especially helpful reading The Goal is the idea of bottlenecks. I wrote about bottlenecks in the last two Coaching Letters: CL #227 about rowing, and #228 on the Strait of Hormuz. In Goldratt’s framing, every system has at least one bottleneck (a constraint that holds back the system as a whole), and the overall performance of the system is limited by that constraint. So improvement efforts in other parts of the system may improve those parts of the system but will not improve the system overall. Goldratt advocates, not surprisingly, paying a great deal of attention to bottlenecks—because knowing about them is just the beginning. Specifically, here’s what you should do:
- Identify the constraint. What is it?
- Exploit the constraint. Maximize use of that bottleneck without throwing more resources at it. For example, if teacher planning time is the constraint, how do you make every minute count?
- Subordinate everything else to the constraint. Align other systems (data, schedules, meetings, leadership) to support that bottleneck.
- Elevate the constraint. Once you’re maximizing the current constraint, consider adding capacity—better curriculum, more targeted coaching, or clearer goals.
- Return to step 1. If you alleviate one constraint, then the bottleneck may shift somewhere else in the system.
I’m just going to repeat one point because it’s a mission-critical idea: improvement efforts directed at parts of the system other than the bottleneck may improve those parts of the system but will not improve the system overall.
I think it’s really important to get our heads around that. I have been in education a really long time, and in that time I have seen schools, districts, states, and the federal government cycle through lots of different theories of improvement and improvement efforts. I am totally confident that the majority of these have not addressed bottlenecks. For example, over the last 15 years, districts have spent hundreds of millions of dollars on technological innovations such as SmartBoards and Chromebooks. And that may have been a good thing to do. But in most cases availability of technology was not a bottleneck limiting student learning, and therefore spending on technology did not, on the whole, improve the system nor, pari passu, move it closer to its goal. Similarly, districts in some part of the country spent tens of millions of dollars lowering class size. I’m not saying they should not have done that either. But the benefits of smaller classes accrue mostly to teachers, rather than students. Class size is not a bottleneck limiting student learning (the research on this goes back a long way.) So the experience of teachers and students may have improved, but the overall system was not improved in terms of moving it closer to meeting its goal. In other words, many of the usual levers—technology, teacher evaluation, class size—are highly unlikely to improve learning because they’re not addressing the system’s real constraints.
So what are the bottlenecks when it comes to optimizing a system for OTL? Here’s one of my early attempts to diagram that, based on the research we were reading and the feedback we were getting from teachers about what impeded their efforts to provide OTL:
Specifically, several experiences this year have revealed that a bottleneck in the systems—in many, but not all, districts—is teachers’ lack of access to high quality tasks. This means that the impacts of improvements in other related parts of the system—teachers’ instructional skills, for example—are being constrained by the bottleneck.
But over time I’ve come to think that this is not the deepest constraint. Upstream of time, task and teaching are the beliefs that educators hold about what students are capable of doing. One of the findings in the research reported in Systems for Instructional Improvement (which I’ve written about a lot, including CL #191 and CL #192) is that teachers can improve in two of the three important areas for improving instruction—pedagogical content knowledge and instructional skill—but if their expectations for students are such that they don’t believe that students are capable of benefiting from the resulting improved instructional practices, teachers won’t use them.
Those beliefs shape all three components of OTL. If educators believe that some students are not capable of engaging in challenging work, they are less likely to provide those students with access to high quality, grade-level tasks. They are more likely to reduce the amount of time those students spend with important content, either by slowing down, removing content, or assigning them to less ambitious work. And they are more likely to use forms of teaching that lower cognitive demand: over-scaffolding, excessive explanation, fragmentation, and simplification.
In other words, beliefs are not simply an instructional issue. They are upstream of the entire OTL system. They determine who gets access to time, task, and teaching in the first place. Which means that even if we improve curriculum, instructional skill, or teacher collaboration, those improvements will be constrained if the underlying beliefs remain unchanged. This is why, in the final version of the diagram, beliefs sit above the rest of the system. They are not one more bottleneck among many. They are the constraint that shapes the others.
And at this point the temptation is to make the same mistake again, and to talk about teacher expectations as though there are individual bad actors whose deficit beliefs lead them to make poor instructional decisions. But that too misses the point. Educators are part of a system, and that system encourages—and in some cases all but requires—them to make work easier for the students thought to be least capable. The system supplies lower-level tasks, rewards coverage over thinking, praises “support” when it means reducing cognitive demand, and treats struggle as evidence that the work is inappropriate rather than evidence that students are doing the work of learning.
So much of what I’ve been doing lately involves reframing the role of the teacher from being maximally helpful to maintaining cognitive demand. Reading The Goal made me realize that beliefs are the bottleneck, but that does not mean the solution is to lecture teachers about having higher expectations. You cannot talk someone into a new mental model any more than you can talk someone into being funnier. Beliefs change when people have different experiences.
Which means that if we want educators to believe that all students are capable of ambitious learning, we have to create the conditions under which that becomes visible and undeniable. We have to ensure that teachers have access to high-quality, grade-level, low-floor/high-ceiling tasks. We have to teach instructional practices that maintain cognitive demand rather than reduce it. We have to redesign collaboration so that teachers can see, together, what happens when students who were presumed incapable are given the chance to think. If you try to change beliefs without changing time, task, and teaching, you are likely to produce frustration. But if you change time, task, and teaching in ways that allow educators to experience students differently, the beliefs may begin to shift.
In other words, beliefs may be upstream of the system, but the way to change them is not to start with beliefs. It is to change the experiences from which those beliefs are built. And that, friends, needs to be a future Coaching Letter…
And in the meantime, if I can do anything for you, please don’t hesitate to reach out. Best, Isobel
Isobel Stevenson, PhD PCC
Author of The Coaching Letter
