A Four-Week Writing Plan
Four weeks is enough to move a band if each session changes one thing. Here is the week-by-week structure, the volume, and how to tell whether it is working.
The principle: one change per session
Writing practice fails in a predictable way. A candidate writes twenty answers, receives twenty sets of feedback, and produces the twentieth answer with the same habits as the first, because nothing in the process required a change.
The alternative is to treat each session as a test of one specific change. Write the answer, read the feedback, choose the one thing that recurs, and write the next answer with that single instruction at the top of the page. Volume then compounds instead of repeating.
This plan is built around that. Four weeks, three to four timed tasks a week, and one criterion in focus per week. Roughly fourteen tasks in total, which is far less than most study plans prescribe and considerably more effective, because every one of them is reviewed.
A note on order. The weeks run from the heaviest criteria to the lightest, because that is the order in which the marks sit: Task Fulfilment and Content and Coherence are 55 per cent between them, and they are also the fastest to change.
Week one: coverage and commitment
Focus: Task Fulfilment. Volume: two Task 1 and two Task 2, timed.
The only thing you are measuring this week is whether every bullet is answered with a specific, whether the register matches the reader, and whether Task 2 commits to one option from the first sentence to the last.
Before each session, write the three bullet labels at the top of your page, with every sub-part named. Date/number/price, not details. After each session, point at each bullet and find the sentence that answers it. Score yourself only on coverage: three out of three, or two out of three.
Add one drill this week that is not a full answer. Take ten prompts and write only the four-minute plan for each: register, bullet labels, invented facts. Ten plans is about forty minutes, and it teaches coverage faster than ten full answers would.
By the end of the week the target is four consecutive answers with full coverage and the correct register. Nothing about elegance. If you hit that, week one has done its job.
Week two: development and structure
Focus: Content and Coherence. Volume: two Task 1 and two Task 2, timed.
Now the question is not whether a bullet is answered but whether it is developed. The instruction at the top of the page this week is a single test: every claim must be followed by a mechanism or a consequence.
On Task 1, convert intensifiers into observable facts. This has been very stressful becomes I have stopped using the stove and I empty buckets twice a day. On Task 2, write the words which means after each claim and finish the sentence.
The second thing to fix this week is paragraph structure. One idea per paragraph, one paragraph per bullet, and an opening line for each that connects to the last. After writing, read only your paragraph openings in sequence: if they summarise your argument, the structure is sound.
Also check your two Task 2 reasons against the stakeholder test: do they affect different groups. If both reasons concern the same people, one of them is a restatement, and this is the week to notice it.
Week three: control and accuracy
Focus: Readability. Volume: three Task 1 or Task 2, timed, plus one rewrite session.
This week the instruction is about the checking routine rather than the writing. Stop at minute twenty-four, run the four passes in order, and log what each pass caught.
Read only the verbs on the first pass, and record every tense drift and agreement error you find. After three sessions you will have a list of three or four errors that recur, and those are your personal errors rather than general English problems. Keep that list visible while writing.
On the second pass, split every sentence over about twenty-five words. Keep count of how many you split. Most candidates find two or three per answer, and every one of them was a risk in a criterion that rewards control.
The rewrite session is the important one. Take your weakest answer from the last two weeks and rewrite it at the next band: cut the padding at the top, supply the missing specifics, turn category words into events. Compare the two versions side by side and count the words you actually changed. It is usually fewer than twenty.
Three to four timed tasks a week, every one reviewed, beats daily writing with no review. Volume only compounds when each session tests one specific change from the last.
Week four: precision and pace
Focus: Vocabulary and timing. Volume: three timed tasks plus one full section, both tasks back to back.
Vocabulary comes last because it is the lightest criterion at 20 per cent and because precision is easier to fix once the structure is reliable. The instruction this week: replace every use of problem, situation, issue, thing and good with the specific word for what you mean.
Build by situation rather than by list. Collect the verbs of damage and remedy for complaints, the nouns for who benefits in civic topics, and the phrases for a warm informal opening. Test each one by writing a full sentence with it in a real prompt context; a word you cannot place today will be misused under pressure.
The full section is the most important session of the month. Both tasks back to back, 53 minutes, no break. It is the only way to find out whether your Task 2 is still disciplined after Task 1 has used your concentration, and for many candidates the answer is that Task 2 quietly loses its commitment sentence and its example.
Finish the week by re-reading your first answer from week one next to your last. The difference is usually not vocabulary, which is the point the whole plan is making.
Tracking whether it is working
Keep one line per attempt: the date, the task type, the four criterion scores, and the single instruction you were working on. Nothing else. A page of notes per answer is not sustainable and will be abandoned by week two.
Look at the four criterion scores as separate lines rather than at the overall band. The overall band moves slowly and hides progress; a Task Fulfilment score moving from 7 to 9 while the others hold is exactly the pattern this plan is designed to produce, and it will lift the total by half a point on its own.
Read your instruction column after four attempts. If the same instruction is still there, the change is not happening and the instruction is probably too vague. Write better introductions is not an instruction; first sentence must contain a fact or a decision is.
Finally, be honest about the coverage column. It is the one thing in this plan that is objectively checkable, it is the heaviest available gain, and candidates consistently overestimate it. Three out of three, or two out of three, written down every time.
Track the four criterion scores as separate lines rather than the overall band. A single criterion moving while the others hold is the pattern that shows the plan is working, and the overall number hides it.
Practise it
Five questions on how to structure practice so that it changes something.