Five Minutes of Planning Buys You Fifteen
Four minutes of planning is the difference between an answer you write once and an answer you have to rescue at minute eighteen. Here is what to decide and what to leave alone.
Why planning pays here specifically
Twenty-seven minutes for Task 1 and twenty-six for Task 2 sounds like enough time to recover from a wrong turn. It is not. At roughly 180 words per task, you have about twenty minutes of composition, and abandoning a half-written answer at minute fifteen leaves you writing 120 words in ten minutes with no time to check them.
So planning in CELPIP Writing is not about elegance. It is insurance against reversal. The four minutes buy you one thing: certainty that the answer you start is the answer you can finish.
The second reason is that three of the four criteria are settled before you type. Whether all the bullets are covered, whether the ideas are ordered, whether the register is right, whether you have committed to an option. Content and Coherence at 30 per cent and Task Fulfilment at 25 per cent are largely decided by a plan. Only Readability and Vocabulary are decided by the writing itself.
The four decisions, in order
One: who is reading this. Read the scenario for the recipient and your relationship to them, and write formal or informal at the top of your scratch plan. On Task 2 the equivalent question is which option you are choosing, and it gets the same treatment: one word, decided, not revisited.
Two: what the bullets actually ask. Reduce each bullet to a two-word label. Date/number/price. Damage/discovery. Remedy/deadline. Three labels is the whole plan for coverage, and it is the thing you check against at the end.
Three: which specifics you are inventing. This is the step people skip and it is the one that decides your band. Before writing, choose the date, the amount, the reference number, the name of the item, the deadline. Fix them now so that you are not inventing mid-sentence and contradicting yourself two paragraphs later.
Four: the order. Usually the bullets are already in a sensible order and you should follow them. Where they are not, move the request to the end and the context to the front. Never reorder so much that a reader would have to look back at the prompt to follow you.
What a finished plan looks like
For a Task 1 email to a moving company about a cracked glass cabinet and a missing box of kitchen items, the whole plan fits on one line:
Formal. 14 Mar / MR-40821 / $1,340. Cabinet cracked + kitchen box gone, found unpacking, 2 calls ignored. Want: inspection this week + compensation, reply by Friday.
That is under forty words and it took ninety seconds. It fixes the register, all three bullets, every invented fact and the closing request. Nothing in it is a sentence, because sentences in a plan are words you are writing twice.
For Task 2 the plan is even shorter. Option, two reasons, one example, one concession: Option B library hours. R1 students after school. R2 cheaper, no new building. Eg: community centre booked Tue/Thu evenings. Concede: arena serves hockey but empty midweek.
Two things are deliberately absent from both plans. There is no opening sentence and there is no vocabulary list. Drafting your first sentence in the plan means writing it twice, and choosing words in advance is how ambitious-but-wrong vocabulary gets into an answer.
Planning Task 2 when you have no opinion
The most common way to lose time on Task 2 is to spend it deciding which option you actually believe in. Your opinion is not scored, so that deliberation is worth nothing, and it can cost three or four minutes.
Replace the question. Instead of asking which option is better, ask which option you can give a concrete example for. Concrete means it has a number, a time or an outcome in it. If one option produces an example within thirty seconds and the other does not, you have your answer regardless of what you think.
If both produce an example, take the one where the second reason is clearly different from the first. The reliable test is whether your two reasons affect two different groups of people. A reason about students and a reason about the town's budget are genuinely two reasons. A reason about students and a reason about young people are one reason written twice, and Content and Coherence notices.
Ninety seconds is the budget for this decision. If you are still weighing at ninety seconds, take whichever option is on top and start writing. A committed argument for the weaker option beats an undecided argument for the stronger one every time.
Fix every invented number before you write the first sentence: the date, the reference, the amount, the deadline. Inventing them mid-paragraph is how answers end up saying 14 March in one place and last week in another.
What not to plan
Do not plan sentences. Writing your opening line in the plan and then copying it out is a straight loss of thirty seconds, and it usually produces a stiffer sentence than you would have written naturally.
Do not plan word choices. A list of impressive words prepared in advance is the mechanism by which wrong words get used, because you will try to place them whether or not the sentence wants them. Vocabulary is scored on accuracy and fit, and words chosen before the sentence exists rarely have either.
Do not plan a word count per paragraph. Counting as you go interrupts composition, and the count is on screen anyway. Check it once at the end.
Do not plan the other task. Task 2 is not visible while you are on Task 1, and there is nothing useful to prepare beyond knowing the shape of it.
Training the plan so it survives the exam
A planning routine only helps if it is automatic, and it becomes automatic through repetition on a timer, not through reading about it. Practise the plan on its own: open a prompt, set four minutes, write the one-line plan, stop. Do not write the answer. Ten prompts planned this way in forty minutes teaches more about coverage than two full answers.
When you do write full answers, keep the plan visible and tick each label as you cover it. The habit you are building is looking back at the bullets while writing rather than only at the end, and that is the habit that prevents the most expensive error in the section.
Afterwards, compare the plan to the finished answer. If the answer contains a fact that is not in the plan, ask whether it contradicts anything. If a label in the plan has no matching paragraph, you have just found, in practice, the thing that would have cost you a band on the day.
Give the Task 2 choice a hard ninety-second limit, and decide it by which option you can produce a concrete example for rather than which you agree with. If the timer runs out, take the top option and commit.
Practise it
Five questions on what belongs in a plan, and what a plan is for.