All activities Professional learning
A coach and principal analyze a coaching-time audit together, including the drift that AI and tech rollouts cause, and write a one-page agreement before the coaching cycle begins.
By November, many coaches find their weeks filled with testing coordination, class coverage, and resetting passwords for the latest platform, while coaching cycles quietly stall. This session comes before the TCEA EIIR coaching cycle begins. A coach and a principal read a fictional two-week coaching-time audit, sort where the time went, and name the sources of role drift, including a campus AI rollout. Then they walk through a six-step conversation and write a one-page agreement: the role, the evaluation boundary, what the coach won't take on, a realistic caseload, how coaching logs stay private, and what AI may and may not do with them.
Hook
Ask each coach to write, privately, the share of last week they spent in real coaching with teachers: co-planning, modeling, observing without evaluating, and looking at student work together. Ask principals to write what they believe their coach's number is. Pairs compare. Say: "The gap between those two numbers is what today's agreement is for."
Facilitator noteThe gap usually surprises both people. Keep it curious, not blaming: principals rarely see the requests that pull coaches away.
Explore
Pairs read Handout A and highlight every line in three colors: deep coaching (cycles, co-planning and modeling tied to student work), surface support (quick resource drops, one-off tech fixes), and non-coaching duties (testing, coverage, duty, administrative tasks). Then they answer: "Which requests came from technology or AI initiatives? Which of those needed a coach, and which needed a help desk, a trainer, or a decision?"
Facilitator noteLook for pairs who spot the AI platform rollout as a drift source: password resets and newsletter slides are real work, but not coaching. Also flag the coaching log: it names teachers with notes like "struggling," sits in a folder the admin team can open, and was pasted into an AI tool to summarize.
Model
With a volunteer as principal, model Step 3 from Handout B. As the coach, say: "I want to be sure teachers can take risks with me. If what I see in coaching shows up in evaluations, they'll stop inviting me in. Can we agree that I don't share observation notes or coaching logs for evaluation, and that you'll ask teachers directly about their growth?" Let the volunteer push back ("But I need to know who's struggling"), and respond with what you can share: which teachers are in cycles, the focus, and de-identified patterns in student evidence.
Facilitator noteElena Aguilar makes the point in The Art of Coaching that coaching tied to evaluation undermines teachers' willingness to take risks. Principals often agree once they hear what they will still get: patterns, not names.
Practice
Pairs work through Handout B, three minutes per step, recording agreements in the right-hand column. (1) Gather the evidence: what the audit and last year's student work show. (2) Clarify the role: what coaching means here, in one sentence. (3) Set evaluation boundaries. (4) Name what the coach won't take on, and who will. (5) Set a realistic caseload: how many teachers in cycles at once. (6) Learn the campus before launching anything new: what the coach will listen for in the first weeks. Coaches paraphrase the principal's view at each step before responding: "What I hear you saying is…"
Facilitator noteStep 4 is where AI and tech drift gets settled. A strong agreement names a specific person or channel for platform troubleshooting and says the coach supports AI integration through coaching cycles, not as tech support.
Create
Pairs fill in Handout C from their Handout B notes. Require the coaching log section: where logs live, who can see them, that teacher names and student information never go into an AI tool, and what AI may help with (drafting agendas, turning the coach's own de-identified notes into a to-do list, drafting sentence stems). Both people sign and set a check-in date to rerun the audit.
Facilitator noteAn agreement that fits on one page gets reread. Push pairs to cut anything that isn't a decision.
Debrief
Ask: "What evidence will tell you in November that this agreement is working?" Take answers such as the share of time in cycles, the number of teachers in cycles, and changes in student work from those classrooms. Then ask principals: "What's one request you'll redirect this month?"
Facilitator noteThe agreement is itself a trust-building tool: it sets expectations clearly and gives both people something to follow through on.
The whole session runs on paper: a printed audit, a printed protocol, and a signed one-page agreement. For the real audit afterward, a coach can tally two weeks of a paper planner by hand into the same three colors.
After the session, a coach can export two weeks of calendar entries to a spreadsheet and categorize each as deep coaching, surface support, or non-coaching, which makes the rerun in November quick. Keep the signed agreement in a document only the coach and principal can open. If the coach keeps a digital coaching log, store it where the admin team cannot browse it, and remove teacher names and student details before using any AI tool on it.
Does it need a screen? A spreadsheet makes the time audit faster to repeat and easier to compare over the year, which is what turns the agreement into data. The conversation and the agreement are better on paper, where both people's attention is on each other.
What you should be able to see or collect if it worked.
A coach and principal analyze a coaching-time audit and write a one-page agreement on role, boundaries, caseload, and log privacy.
Run a real two-week audit of your own time using the three colors, bring it to your principal, and sign your agreement. Tell teachers in one sentence what coaching is on your campus and what happens to your notes, then rerun the audit before November.
