Your coaching model is fine. The sludge is killing it.
Most networks don’t need another instructional framework — they need the one they already bought to happen every week.
Buy another coaching model if you want. Most of the ones on the shelf are fine.
That is not the problem.
The problem shows up on a Tuesday afternoon: a coach finishes three walkthroughs, scribbles look-fors on a clipboard or in a notes app, promises a follow-up, and then the week swallows the follow-up. Friday’s PLC is about something else. Next week’s visit starts from memory. The teacher never sees a crisp next step. The model was never the missing ingredient. The operating rhythm was.
We call that administrative sludge — the coordination overhead that stops good instructional intent from becoming weekly practice. It is boring. It is also where coaching dies.
This is an opinion piece about that sludge. It is not a case study, and it is not a claim about any named school.
What sludge looks like in real life
- Observation notes living in five places (email, Drive, clipboard, chat, someone’s head).
- Look-fors that exist in a binder nobody opens during the visit.
- “Feedback” that is a hallway compliment, not a recorded focus for next time.
- Coaches spending more time hunting forms than talking about teaching.
- Leaders asking for coaching data and getting a scramble instead of a trail.
None of that is fixed by a new workshop on questioning techniques.
The pitch that actually matches the work
A useful K–12 line is blunt: you already have an instructional model — this is what makes it happen every week.
That sentence is about follow-through, not novelty. Bring your model. Keep your language. Keep your look-fors. What you need is a shared place where:
- What was observed can be found later.
- The instructional focus is named so the next visit is not a reset.
- The next step is visible to coach and teacher.
That is coaching support, not evaluation theater, and not “AI will coach for you.”
Why AI made the sludge worse before it helps
Teacher-facing AI can draft a beautiful paragraph about a lesson the coach half-remembered. Curriculum vendors are baking AI into materials schools already buy. None of that automatically creates a weekly coaching cycle.
If anything, faster drafting can hide the sludge: the write-up looks done while the practice never changed. Policy PDFs pile up. Approved-tool lists grow. Meanwhile the coach still cannot find last week’s focus.
Tools that help are the ones that reduce coordination cost — capture, focus, next step — without pretending judgment is optional.
What to stop buying
- Another framework that assumes unlimited coach time.
- A dashboard nobody opens between board meetings.
- Student-facing “innovation” while adult practice has no weekly rhythm.
- Case-study theater that substitutes logos for operating detail.
What to build toward instead
Ask boring questions in the fit conversation:
- Who enters observations, and who reads them?
- How does a look-for become next week’s visit plan?
- What happens when a coach is out for two weeks — does the trail survive?
- Are coaching notes separated from HR evaluation on purpose?
If those answers are fuzzy, software and advisory work should start there — not with a rebrand of your instructional model.
CoachingOS, as we build it, is aimed at that cycle: observations, focus, and next steps for teams that already know how they want to coach. Product description, not proof. Your campus still has to run the conversations.
If the sludge is the thing your network feels every week, talk with us about coaching-support practice — software, custom build, or advisory — without the theater.