How to Make the Most of an IELTS Writing Bank: A Practice System for Both Tasks
A writing bank is a collection of practice prompts and model answers covering both IELTS Writing tasks: Task 1 charts, processes and letters, and Task 2 essays. Worth saying plainly: everything in a legitimate writing bank — including everything on this site — is original material written for practice. Real exam papers are confidential and never published, and a writing bank's value has nothing to do with predicting your test question. Its value is that Writing is the skill where structured comparison against a standard produces the fastest measurable gains — if you use the bank as a training system rather than a reading list. This guide is that system, task by task.
The one rule that makes everything else work
Write before you read. Every study session starts with your own timed attempt — 20 minutes and 150+ words for Task 1, 40 minutes and 250+ words for Task 2 — before you open any model. Reading a model first feels efficient and teaches almost nothing, because you never committed to your own decisions and so the comparison has nothing to bite on. Writing first converts the model from entertainment into feedback.
Using the bank for Academic Task 1
When you compare your chart, graph or process description against a model, audit four specific decisions:
- The overview. Does the model's overview capture the two or three dominant patterns without a single statistic? Does yours exist at all? A missing or buried overview is the single most common Task 1 band-capper.
- Data selection. Count the figures the model actually quotes — typically far fewer than candidates expect. Models group and summarise; weak answers narrate every number in order. If your response reads like a data table transcribed, you have found your gap.
- Grouping logic. How did the model divide the information between its body paragraphs — rising vs falling, before vs after, category vs category? Naming its grouping principle teaches you to find one in your own next attempt.
- Comparison language. Harvest the precise machinery: "more than doubled," "narrowed steadily," "peaked at," "remained broadly stable." These small patterns transfer directly; whole sentences do not. Our guide to describing charts in Task 1 covers the full toolkit, and process diagrams have their own dedicated method.
Using the bank for General Training letters
Letter models earn their keep on two axes. First, bullet coverage: the task's three bullets must each be developed, not merely mentioned — check how many sentences the model spends per bullet and compare your ratio. Second, tone control: study how the same request changes clothing across formal, semi-formal and informal letters ("I would be grateful if you could…" versus "Any chance you could…?"), because consistent register is what separates Band 6 letters from Band 7 ones. Collect openings, closings and request patterns as a tone toolkit rather than memorising any whole letter — the full framework is in our guide to General Training Task 1 letters.
Using the bank for Task 2 essays
Compare your essay against the model criterion by criterion, in this order:
- Task Response: put the question next to both essays and check every part of it is answered — both views discussed, opinion given, both halves of a two-part question addressed. Note how the model's position appears in the introduction and never wavers.
- Structure: label what each of the model's paragraphs does, then label yours. If any of your paragraphs needs a label like "more ideas," you have found the problem — one developed idea per paragraph is the pattern, as our Band 7 structure guide lays out.
- Development: watch how the model extends one idea across four or five sentences — claim, explanation, example, consequence — where weak essays state three ideas in three sentences and move on.
- Language: harvest at most three collocations and one grammatical pattern per model. Never lift introductions or conclusions wholesale; examiners recognise prefabricated content, and it fails the exact criterion memorisation is meant to help.
Close the loop with objective feedback
Self-comparison has a ceiling: writers systematically forgive their own errors and cannot reliably see which criterion caps their band. After the compare-and-rewrite step, get your essay scored against the actual criteria — our IELTS writing checker returns criterion-level bands and specific fixes, which turns "I think it's better now" into data. Then schedule a fresh prompt from the Task 2 topic bank a few days later and check whether the improvement survives on unseen material. That cycle — timed attempt, model comparison, targeted rewrite, objective score, fresh retest — is the entire method; a bank of original practice prompts is simply what keeps the cycle fed. One fully processed prompt per task per week compounds into a visibly different writer within two months.
What to skip
Skip phrase-list hoarding ("sophisticated linkers" pasted into every essay read as mechanical, not advanced), skip rewriting the same essay five times (the exam tests transfer to new prompts, so practise transfer), and skip any resource promising real exam content — it does not exist publicly, and time spent chasing it is time taken directly from the practice that works.
Track the trend, not the essay
Finally, keep a simple log: date, prompt, task type, criterion scores, and the one error you are currently hunting. Individual essays fluctuate — a familiar topic flatters you, a hard chart humbles you — but a six-week trend line across unseen prompts is honest. When the log shows your capping criterion has shifted (from grammar to lexis, say), shift the training with it. That is the whole discipline: the bank supplies the volume, the log supplies the direction.