Discover tools, strategies, and resources for instructional coaching. Explore insights to support teachers and enhance professional learning.
In August, your campus leadership team launched something. A new instructional framework, an AI pilot, or a revised coaching cycle. By October, what is actually happening in your building looks different from what you launched, and nobody ever made a decision to change it.
That gap is usually not a failure of leadership commitment. It is the result of patterns that operate on your campus whether or not you account for them. Five of those patterns are named below, illustrated through Lincoln Middle School, a composite campus. What they give you is shared language for naming a pattern out loud in a meeting, which is often the first step toward fixing it.
Goodhart’s Law: The Number Gets Healthy, The Practice Does Not
Lincoln’s leadership team set a walkthrough goal in August: 25 classroom walkthroughs per administrator per month. By October, every administrator was hitting 25. To get there, the walkthroughs got shorter, clustered in the periods that were easiest to protect, and landed most often in classrooms where a five-minute visit would go smoothly.
When a measure becomes a target, it stops being a good measure.
Nobody cheated. The number is real and the team reports it with confidence. It also no longer tells anyone what instruction looks like at Lincoln, because the goal quietly reshaped the behavior it was built to observe. The metric did not fail. It succeeded, and now leadership has a clean report with no real read on what is happening in classrooms.
Hanlon’s Razor: Overload Looks a Lot Like Resistance
Never attribute to resistance what circumstance already explains.
A coach at Lincoln has a teacher who skipped two coaching cycles and left two emails unanswered. The coach reads it as pushback on the new framework and stops putting him on the schedule. The actual cause was a September section change that added a fourth prep, arriving the same month a parent went into the hospital.
The cost is uneven. If you assume overload and you are wrong, you lose a little time and try again in two weeks. If you assume resistance and you are wrong, you end a working relationship that was never actually broken, and you rarely find out.
One firm limit applies here. This is a lens for extending charity, never a tool for reclassifying a legitimate objection as fatigue. If a teacher tells you the framework does not work in her content area, that is an argument you owe an answer, not a symptom you manage.
The Pareto Principle: Most of Your Hours Go to a Few of Your People
Roughly 80 percent of effects come from 20 percent of causes.
An instructional coach at Lincoln supports 22 teachers. Four of them absorb most of her coaching hours. That concentration is normal, and by itself it is not a problem.
The problem is what the split does not tell you. Knowing that most of your time goes to a few people says nothing about whether that time is landing. Teachers who need the most support are often least likely to convert coaching hours into changed practice, because what limits them is workload, placement, or fit, and more coaching resolves none of the three. Meanwhile, 18 teachers who would move quickly after two focused conversations get a check-in in the hallway.
Here is a move you can make this week. Pull your last month of calendar entries, sort them by teacher, and for each of your top three or four time consumers, name one thing that changed in their practice. If you cannot name it, the answer is not more hours.
The Law of the Instrument: When the New Tool Becomes the Only Answer
When the only tool you have is a hammer, every problem starts to look like a nail.
Lincoln piloted an AI writing feedback tool in English classes last spring, and teachers liked the results enough that the tool became the default answer to almost anything touching writing. By October, a science teacher is running lab reports through it for feedback on data analysis, a task it was never built to evaluate, and a counselor is running discipline narratives through it looking for behavior patterns, which is closer to data analysis than writing feedback. The tool did exactly what it was designed to do. The problems it got assigned to were not.
Nobody misused the tool on purpose. It was available, it was already trusted, and it was easier to reach for than to ask what the problem actually needed. A familiar tool applied to the wrong problem does not usually fail loudly. It produces output mediocre enough to accept and specific enough to feel like an answer, and the actual problem, whether a science rubric or a behavior pattern, goes unexamined because something already got done.
Gall’s Law: A Full Launch Leaves You Nothing To Compare
A complex system that works has evolved from a simple system that worked.
Lincoln’s August launch shipped complete. The framework, the observation rubric, and the data protocol all went live across every grade level in the same week. By October the results are mixed, with sixth grade strong and eighth grade stalled.
The cost is not risk, it is evidence. Because everything started at once, you have nothing to compare against. You cannot tell whether eighth grade stalled because of the rubric, the data protocol, the master schedule, or the framework itself.
None of that is a verdict on August. It is guidance for the next launch, and there is something you can still do with this one. Choose a single component and change it in a single grade level for six weeks. That gives you one clean comparison, which is one more than you have today.
An Honest Caveat
Their value is speed. Saying “that might be Goodhart” in a meeting takes four seconds and puts a pattern on the table that would otherwise take much longer to explain from scratch.
These are not laws in the way that gravity is a law. None of them is falsifiable, and you can find a campus that contradicts each one. Treat them as vocabulary, not evidence.
Run This In One Meeting
Pick one initiative that is live on your campus right now. Bring these five questions to a single leadership team or coaching meeting and answer all five about that one initiative before you try to solve anything. Budget about 25 minutes.
Copy this script
These same patterns show up at the district level, where the timelines are longer and the dollars are larger, and they look noticeably different from that seat.
