If you are standing up an instructional coaching program, you are making one of the highest-leverage moves available to a school or district. Coaching is where professional growth becomes personal: a teacher and a coach working a real problem in a real classroom, week after week. It is also one of the most predictable ways to keep good teachers, because professional growth and the sense of being supported are exactly the conditions that make people stay. Stanford's Center for Education Policy Analysis finds that professional growth and administrative support are among the strongest predictors of whether teachers stay. A coaching program, done well, is how you deliver both.
Done poorly, coaching becomes another initiative that fades by November. The pattern is familiar: a coach gets hired, gets pulled into testing coordination and substitute duty, never protects real time, and by spring no one can say what changed. This guide is the antidote. It walks through the decisions that determine whether a coaching program lands, in the order you should make them, and it is honest about the failure modes that quietly kill these programs.
The through-line of this guide
A coaching program is not a job posting. It is a set of design choices: a clear purpose, a model, a repeatable cycle, protected time, a trust boundary with evaluation, and a way to see whether it is working. Get those right on a small scale first, then grow.
Why coaching, and why it is worth doing right
Most professional development is an event. Teachers attend, nod, and return to classrooms where nothing structurally changes. Coaching is different because it is job-embedded and ongoing. It meets a teacher inside the actual work, targets one thing at a time, and comes back to check. That is why it moves practice when a workshop does not.
It is also a retention lever. Teachers who feel they are growing and supported are far more likely to stay, and coaching is one of the clearest signals a district can send that it is investing in its people rather than just evaluating them. But the payoff only arrives if the program survives contact with the school calendar. The rest of this guide is about making sure it does.
1. Define the purpose and the outcomes
Before you hire anyone or pick a model, answer one question in a single sentence: what is this coaching program for? A program built to support new teachers looks nothing like one built to roll out a new literacy curriculum, which looks nothing like one built to lift a few specific instructional practices across every classroom. If you cannot name the purpose, coaches will invent their own, and the work will scatter.
- Write the purpose as one sentence, then name two or three outcomes you would actually check at the end of the year.
- Decide who the program is for first: new teachers, a subject area, a school, or a district-wide priority. Do not try to serve everyone in year one.
- Connect coaching to goals teachers already have, so it reinforces the year's priorities instead of adding a parallel one.
Tie coaching to the goals it should move: goals and SLOs.
2. Choose an instructional coaching model
There are several well-established families of coaching, and you do not need to invent one. You need to choose the family that fits your purpose and be consistent about it. In broad strokes, most instructional coaching models fall into a few types:
- Goal-directed, teacher-led coaching. The teacher sets a goal and the coach partners toward it. Strong for buy-in and for experienced staff who know what they want to improve.
- Directive, practice-based coaching. The coach models a specific technique and the teacher rehearses it with focused feedback. Strong for new teachers and for rolling out a defined set of moves.
- Content or curriculum coaching. The coach is grounded in a subject or a new program and supports teachers implementing it. Strong when a specific adoption is the reason coaching exists.
- Peer or team coaching. Teachers coach each other in structured cycles, often facilitated. Strong for building capacity and culture when dedicated coaches are scarce.
Most real programs blend these, for example directive practice for first-year teachers and goal-directed partnership for veterans. The point is to choose on purpose rather than let each coach improvise a different approach. Write down the stance you expect, so a new coach can learn it.
3. Design a repeatable coaching cycle
The engine of any coaching program is the cycle: the short, repeating loop that turns a conversation into a change in practice. Whatever model you chose, the cycle gives it structure. A workable cycle is small on purpose, focused on one or two moves rather than a whole rubric, and it comes back around to look again.
- Agree on one goal. Coach and teacher name a single, concrete focus for the cycle.
- Gather evidence. The coach observes, co-teaches, or reviews artifacts to see the goal in action.
- Reflect and plan. They meet, look at what happened, and choose the next move together.
- Try and look again. The teacher tries it, and the cycle repeats until the goal is met.
Decide the cadence up front. A cycle that runs a few weeks and meets on a predictable rhythm is far more durable than open-ended coaching that depends on whoever has time. Give coaches a simple, shared place to log cycles so the work is visible and does not live in one person's notebook.
See how cycles are structured and logged: instructional coaching, with classroom evidence from walkthroughs and observations.
4. Decide coach roles and caseloads
A coach with no defined role and an unbounded caseload is a coach who will end up doing everything except coaching. Be concrete about what the role is, what it is not, and how many teachers one coach can realistically serve well. A coach spread across too many teachers cannot run true cycles with any of them.
- Write a clear role description that states coaching is the primary job, not a collection of leftover duties.
- Set a realistic caseload based on how many active cycles a coach can carry at once, and resist the urge to inflate it.
- Name what a coach does not do: not the evaluator, not the substitute pool, not the testing coordinator.
- Support your coaches too. They need their own learning, a community of practice, and a manager who protects the role.
Coaches are educators with growth needs of their own: PD and license tracking.
5. Protect the time
This is the step programs skip, and it is the one that kills them. Coaching time is the first thing sacrificed when a school is short-staffed, and a coach who covers classes all week is not coaching, no matter what the org chart says. If the time is not protected in writing and defended by leadership, the program will quietly stop happening.
- Put coaching time on the master schedule as a protected block, not as whatever is left over.
- Agree in advance on the rules for when a coach can and cannot be pulled for coverage, and hold to them.
- Give teachers protected time to be coached as well, so the whole burden does not fall on their prep periods.
- Make coaching time visible on the same calendar as everything else, so leaders can see when it is being eroded.
Coaching sessions belong on the shared calendar alongside every other dated commitment, so they are planned rather than squeezed in.
6. Build trust: keep coaching separate from evaluation
Coaching only works when teachers can be honest about what they are struggling with. The moment they suspect that what they say to a coach could show up in a rating, they stop being honest and start performing. That is why the single most important cultural decision in launching a program is making the boundary between coaching and evaluation explicit and real.
- State plainly, up front, that coaching is developmental and confidential, and that it does not feed into a teacher's formal evaluation.
- Where staffing allows, keep the coach and the evaluator as different people. Where it does not, separate the conversations clearly and name which is which.
- Keep coaching notes out of the evaluation record, and be transparent about who can see what.
This distinction is subtle enough that it deserves its own treatment, because getting it wrong undoes everything else. If coaching starts to feel like evaluation, teachers disengage and the program becomes theater.
Go deeper on the boundary: instructional coaching vs. evaluation, and see how the tools stay separate across coaching and evaluations.
7. Measure whether it is working
A program you cannot see is a program you cannot defend at budget time, and one you cannot improve. You do not need a research study. You need a simple, honest picture on three levels: is the coaching happening, is practice changing, and is it moving the outcomes you named in step one.
- Activity. Are cycles actually happening, and how often? A count of active and completed cycles tells you whether the program is real or aspirational.
- Practice. Is teaching changing? Look for the targeted moves showing up in classrooms over time through walkthroughs and coach reflections.
- Outcomes. Is it moving what you set out to move, whether that is new-teacher retention, a curriculum rollout, or a specific instructional priority?
Pair the numbers with the human read. Ask coached teachers whether the support was useful through short surveys, because a program can be busy and still not be helping. Measurement is not about proving coaches are productive. It is about learning fast enough to improve the program while it is still young.
Gather the signal that tells you if coaching is landing: staff surveys.
The failure modes to name before you launch
Coaching programs rarely fail loudly. They fade. If you name the common failure modes out loud at launch, you can design against each one instead of discovering them in the spring.
The coach becomes a floating substitute, testing lead, and hall monitor. Guard against it with a written role and protected time that leadership actually defends.
Coaching happens on whatever minutes are left, which is none. Put cycles on the master schedule and set the coverage rules in advance.
Teachers stop being honest and start performing. Keep the boundary explicit, keep notes out of the rating, and separate the roles where you can.
There are quieter ones too: a program with no purpose that drifts, caseloads so large that no real cycle ever finishes, and a program no one measures, so no one can tell whether it earned its budget. Every one of these is a design choice you can make differently, which is the good news. None of them is bad luck.
Coaching is one connected part of the educator experience
Notice how much of launching a coaching program depends on things outside coaching itself. It draws on the goals teachers set, the classroom evidence from walkthroughs, the calendar that protects the time, the clean line to evaluation, and the surveys that tell you if it is helping. A coaching program run in isolation from all of that is exactly the one that fades.
Coaching lands when it is one connected part of the larger educator experience, from the day a teacher is hired to the day they are thriving. Onboarding hands off to coaching. Coaching feeds the growth goals. Walkthroughs supply the evidence. Surveys close the loop. When those share one picture of each teacher, coaching stops being a standalone initiative and becomes part of how the district develops people.
That is the problem EX in Education was built to solve. It runs the connected educator experience start to finish, so a coaching program shares the same ground as onboarding, growth goals, observations, and surveys instead of standing alone. Bring your own coaching model, or start from a proven template, and run it across every school so every educator moves from hired to thriving.
Frequently asked questions
How do you start an instructional coaching program?
Start by defining the purpose and the outcomes you want, then choose a coaching model and a repeatable cycle to match. Decide who coaches, set realistic caseloads, and protect dedicated time on the calendar. Make it explicit that coaching is developmental and separate from evaluation, launch with a small group first, and measure whether practice is actually changing before you scale.
What does an instructional coach do?
An instructional coach partners with teachers to improve their practice through cycles of observation, modeling, planning, and feedback. Coaches co-plan lessons, gather classroom evidence, help teachers set and pursue a specific goal, and support new initiatives. The role is developmental and confidential, which means it works best when it is kept distinct from the person who assigns a formal rating.
What is an instructional coaching cycle?
A coaching cycle is a short, repeating loop of focused improvement. A coach and teacher agree on one concrete goal, gather evidence from the classroom, meet to reflect and plan a next move, and then try it and look again. Cycles usually run a few weeks and concentrate on one or two moves at a time rather than an entire rubric, so change is visible and manageable.
How do you measure instructional coaching?
Measure coaching on three levels: activity, practice, and outcomes. Track whether cycles are actually happening and how often, look for observable changes in teaching over time, and connect the work to the goals coaching was meant to move, such as retention or a specific instructional priority. Pair the numbers with teacher feedback so you learn whether the support feels useful, not just whether it occurred.