Use a photo of your whiteboard to plan tomorrow's opener

Use a photo of your whiteboard to plan tomorrow's opener

The whiteboard at 3.05pm is the most honest record of what actually happened in that lesson, and it gets wiped in under a minute. Snap it before you clean it.

Wednesday of Term 3 Week 1. You have taught two days, you know the class in a rough way, and Thursday's opener still lives in the carpark drive home. The problem with the carpark plan is not the drive, it is the source. You are planning from memory of the lesson, and memory of a lesson at 3.30pm is already a summary of a summary. The board is not a summary. The board has the worked examples, the student contributions, the errors that got called out, the questions that got left half-answered. Feed it to Claude or ChatGPT with three lines of context and you have Thursday's opener drafted before you get to the staff room.

Most teachers have never used the photo input on their AI of choice for lesson work. It is a habit worth building.

The prompt

[Attach: photo of the end-of-lesson whiteboard, taken before wiping.]

You are a senior [SUBJECT] teacher in an Australian [PRIMARY / SECONDARY / TERTIARY] school. I am about to plan the opener for the next lesson with this class. I have attached a photo of the whiteboard at the end of today's lesson.

CONTEXT (three lines, that is all)
- Year level: [YEAR LEVEL, e.g. Year 9]
- What we were meant to cover today: [ONE SENTENCE, e.g. "quadratic factorisation, extracting a common factor first"]
- What actually got done: [ONE SENTENCE. Honest. If half the class was still stuck at question 3, say so.]
- Next lesson is in [HOURS OR DAYS, e.g. "17 hours"] and lasts [MINUTES, e.g. 50] minutes. It needs to move us toward [ONE SENTENCE about the unit destination, e.g. "solving quadratics by factorising, ready for the summative in week 4"].

DO THIS
1. Read the whiteboard. Tell me in three lines what you can see the class actually worked on, including any student contributions, common answers, or errors that are visible.
2. Name the one thing the board suggests the class is still shaky on. If the board does not give you enough to say, say so. Do not guess.
3. Draft a five-to-eight minute opener for the next lesson that (a) retrieves the shaky thing from today, and (b) sets up the next content. Include the exact wording of the question or task I should put on the board, and the two-line teacher script that goes with it.
4. Flag one thing about my read of the lesson (from the CONTEXT above) that is not backed by the board, so I can check whether my memory of the lesson matches what actually happened.

Keep the whole answer under 300 words. This is a planning prompt, not a lesson plan.

What it does well

The photo gives the model something you would not otherwise think to type out: the actual state of the class at the end of the lesson, including the mistakes and contributions your carpark memory has already smoothed over. Turning that into a targeted opener in under a minute is exactly the workflow that dies when you plan from memory. Step 4 (flagging where your read is not backed by the board) is quietly the best bit; it catches the "I think they got it" that a class did not get.

Where it falls over

If the board is a jumble of shorthand, wiped half-off, or layered from three lessons, the model is guessing and its guesses will be plausible-sounding rubbish. It cannot tell what happened between the writing on the board (only what got written), it does not know which two students were arguing about question 4, and it cannot see the pile of exit tickets on your desk. Treat the output as a starter, not a plan. If you teach in a room where the "whiteboard" is a smart board that wipes itself, screenshot the last state of it before you close the file.

Try it like this

• Primary maths: photograph the anchor chart plus student worked examples for column addition. The model catches which students still line up digits wrong.
• VCE English: photograph the annotated quote wall you built during a text study. The model drafts a retrieval Do Now on which quote linked to which theme.
• If you only have thirty seconds: skip step 4 and just ask for the opener. The board plus the year level is enough for a decent starter, and thirty seconds beats the carpark plan every time.

Run it every day for a week and see whether the Thursday openers land differently than the ones you wrote from memory. The photo takes three seconds. The opener drafts in under a minute. That is the whole ask.