What a Coaching Cycle Actually Looks Like, Week by Week

Why it has to be a cycle at all

In 1981, Bruce Joyce and Beverly Showers published a paper about why teacher training kept not working. They broke training into its parts: study the underlying theory, watch the strategy demonstrated several times, then practice it with feedback. Reasonable. It is still what most professional development looks like on a Tuesday in August.

Their finding was that those three parts produce a teacher who can perform the skill, and not a teacher who uses it. Getting it into what they called the active repertoire was new learning on top of the training, and it required a further step: on-site coaching.^[1]

Bound that properly. It is 1981, the studies behind it were small, and professional development was a different animal. Nobody should quote it as a modern effect size and we are not going to.

What survives the forty years is the distinction, and it is the useful part. Knowing how to do something and doing it in front of twenty-two children on a Tuesday in October are two different accomplishments, and only one of them gets built in a training room. We wrote about the personal version of that gap in You Might Be Teaching It Right. You'd Never Know., and about what happens when there is nobody in the building whose job is to close it in What to Do When Your School Has No Instructional Coach.

A cycle is just the shape that closes that gap. Four weeks is the common version. Some run three, some run six. The number matters much less than whether the last week happens.

Week one: name one thing, and write it down

The first conversation has one job. Leave it with a single thing to look at, stated so plainly that a stranger could count it.

The failure mode here is generosity. A coach who cares will leave week one with three goals, because all three came up and all three are real. Three goals is zero goals. The teacher will do a thin version of each and none of them will take.

Ask something like this.

"What is one thing you want your students doing that they are not doing yet?"

Not what the teacher wants to be better at. What the students would be doing differently if it worked. That single reframe does most of the work, because it moves the target off the adult's performance and onto something observable in the room.

Then make it countable. This is the part that gets skipped and it is the part that makes week two possible.

• Not workable: "improve student engagement."
• Workable: "more than four different students offer an idea without being called on."
• Not workable: "better questioning."
• Workable: "after a student answers, I ask a follow-up before moving to the next child."

Write it down. On paper, in an email, anywhere both people can see it in three weeks. An unwritten goal quietly becomes whatever the observer noticed instead.

For a new teacher, or in the first weeks of a year, the honest version of week one is usually narrower than the coach wants it to be. We made that case in The First Weeks of School Set the Whole Year.

Week two: watch, and describe without a verdict

Here is the research finding that should change how you run this week, and it comes from the same two people.

Showers and Joyce spent more than a decade building peer coaching in schools. In 1996 they wrote up what they had changed along the way, and one item on that list is remarkable. They took verbal feedback out of the coaching process. Not softened it. Removed it.

Their reason was that it would not stay neutral. However they framed it, teachers heard evaluation, because years of clinical supervision had trained everyone to wait for the good news and then the bad news. Peers giving peers feedback slid into supervisory, evaluative comments no matter what the training said.

And then the sentence that matters: omitting feedback did not depress implementation or student growth.^[2]

Bound this one carefully too. That was peer coaching between colleagues with no authority over each other, in the 1990s, and "did not depress" is not the same as "improved." It does not prove feedback is worthless in every setting.

What it does establish is narrower and more useful. The evaluative half was never the active ingredient. Two teachers watching each other and describing what happened carried the gains on its own. That is the empirical version of the argument we made in Coaching vs. Evaluation, and it is why the distinction is a practice rather than a philosophy.

So in week two the observer collects the countable thing from week one and very little else. Times. Counts. Direct quotes.

• "9:14, you asked the question. 9:15, you called on Marcus."
• "Four students offered an idea unprompted. Three of them were at the back table."
• "You said 'good' eleven times and 'say more about that' twice."

No adjectives. No "the energy was great." A teacher can argue with an adjective and cannot argue with a tally, and the point is not to win, it is to have something in front of both of you that neither of you has to defend.

The teacher looks at that record first and says what they notice. Not the coach. This is the same muscle we described in Reflection Is the Most Underrated Teaching Skill, and week two is where a cycle either builds it or replaces it.

Week three: one move, small enough to try Tuesday

Now the cycle has to produce something to do.

Jim Knight's Impact Cycle, which is the version most instructional coaches in the United States have been trained on, runs identify, learn, improve.^[3] Week three is where "learn" turns into "improve," and the whole thing lives or dies on how small the move is.

Use this test. Can the teacher attempt it in the next five days without changing their schedule, their seating, or their unit plan?

If it needs a new curriculum, it is not a coaching move. If it needs approval, it is not a coaching move. If the teacher has to remember it at 7:40 a.m. on Monday and it involves four steps, it is not a coaching move.

• Too big: "build more student discourse into your math block."
• Right size: "after the first student answers, count to three before you say anything."
• Too big: "improve your transitions."
• Right size: "give the two-minute warning before cleanup on Tuesday and Thursday, and notice what changes."

The coach can put two or three options on the table. The teacher picks. That ordering is not a courtesy, it is the mechanism. A move the teacher chose gets tried on a bad week. A move the coach assigned gets tried when someone is watching.

Decide the date of week four before this conversation ends. Not "let's touch base." A day, on both calendars, before either person stands up.

Week four: come back

This is the one. Everything above it is an observation, and observations are cheap.

Week four is also the week that quietly stops happening, in almost every school, almost every year. Nobody decides to skip it. A fire drill, a parent meeting, a stomach bug, a testing window, and by the time anyone looks up it is November and the goal from week one is a piece of paper in a drawer. We described that arithmetic from inside a building in The Quiet Problem in Private Schools and measured it across programs in Instructional Coaching for Small Schools. It is not a caring problem. It is a capacity problem, and it is the most predictable failure in this entire process.

The conversation itself is short. Two questions.

"Did you try it?"

Ask it flatly, and make it genuinely safe to say no. If the answer is no, the interesting information is why, and it is usually that the move was too big, which is a week three problem and not a character problem. Fix the size and run it again.

"What happened?"

Not "did it work." What happened. Then go back to the countable thing from week one and look at it together.

Then one of three things is true. It worked and it sticks, so pick the next thing. It worked and it is fragile, so run the same goal one more cycle. It did not work, so the goal was wrong or the move was wrong, and you now know something you could not have known in week one.

All three of those outcomes are a successful cycle. Only skipping the week is a failure.

Four meetings is not a cycle

The word cycle is doing real work. What separates this from four meetings is that week four becomes week one.

The goal either graduates or repeats. The record from the last round is the starting evidence for the next one. By the third loop the coach is not asking what the teacher wants to work on from a blank page, they are reading a history, and the conversation starts somewhere specific.

That accumulation is the actual product. One cycle is a nice conversation. Six cycles is a teacher who has changed, and a record of how.

Where it actually breaks

Three failures, in the order we see them.

Week four is never scheduled during week three. Fixable in ten seconds and almost never done. If the date is not on two calendars before people stand up, it will not happen. This is the single highest-return habit in the whole process.

The person coaching you is the person who evaluates you. In a small school this is often unavoidable, and pretending otherwise helps nobody. It can be worked with, but only if it is said out loud in week one, including what will and will not travel from these conversations into a review.

There is more than one cycle running at once. A teacher can be inside a literacy consultant's cycle, a curriculum vendor's training, and a building walkthrough process in the same month, each with its own goal and its own follow-up. Nobody in that picture did anything wrong. But four goals arrive as no goal, because the rational response to four watchers is a thin version of all four, and thin versions do not take. That is a coherence problem, it belongs to whoever bought the programs rather than to the teacher living underneath them, and it is worth its own article.

How eduimpact fits, honestly

We covered this in detail last week in Can AI Coach a Teacher?, and the answer has not changed in seven days.

Map it onto the four weeks. Week one, deciding what matters for this teacher this year, is human, and it stays that way. Week three, choosing the move, is mostly human, because the person who knows you picks the one you will actually do.

Weeks two and four are different. Watching and describing without a verdict is something a system can do accurately and without history in the room. And coming back is a scheduling and memory problem, which is exactly the sort of thing that should never have depended on a person having a free Thursday.

That is the claim, and it is deliberately a small one. eduimpact extends the reach of a coaching cycle into the weeks a human coach cannot physically get to. It is not the coach, it does not decide what you work on, and a person reviews what it produces before it reaches you. Nothing is rated. Nothing is ranked.

If your school has an instructional coach, this gives them more weeks. If your school has nobody in that role, it gives you a description of your own teaching that is better than memory, which is more than most teachers have ever been handed.