What a Band 9 IELTS Essay Looks Like, and Why Most Advice Is Wrong

Search for band 9 essays and you will find hundreds. Read ten of them and you will notice they all sound the same: long introductions, heavy linking words, vocabulary that is clearly there to be noticed. Most of them are not band 9. They are competent band 7 essays with a 9 written at the top by whoever posted them.

This matters because you copy what you read. If you model your writing on a band 7 essay labelled 9, you practise your way to 7 and cannot understand why.

Here is what the four public band descriptors actually ask for at band 9, what that looks like in real sentences, and how to tell a genuine band 9 paragraph from an imitation.

What band 9 says in the descriptors

The band descriptors are published by IELTS and describe each criterion at each band. At band 9 they are short, and their shortness is the point.

Two phrases in that table do most of the work.

The first is "attracts no attention" under Coherence and Cohesion. Linking words that you notice are, by definition, attracting attention. An essay stacked with Firstly, Moreover, In addition and Last but not least is not scoring 9 on that criterion, no matter how correct each one is. At band 9 the reader follows the argument without ever thinking about the joins.

The second is "rare minor errors occur only as slips". A slip is the kind of mistake a confident native writer makes when tired: a missing plural, a mistyped article. It is not a pattern. If the same error appears three times, it is not a slip, and Grammatical Range and Accuracy will not sit at 9.

A band 7 paragraph and a band 9 paragraph on the same question

The question: Some people believe that governments should spend money on public transport rather than on building new roads. To what extent do you agree or disagree?

Here is a body paragraph at roughly band 7.

Firstly, public transport is very beneficial for the environment. Cars produce a lot of pollution and this is bad for the air quality in big cities. Moreover, if the government invests in buses and trains, many people will use them instead of their cars. For example, in my country many people use the metro every day. Therefore, pollution will be reduced significantly.

There is nothing wrong with it. The grammar is accurate, the idea is relevant, the paragraph is organised. It sits at 7 because of three specific things.

Now the same idea at roughly band 9.

The environmental case is the strongest one. A bus carrying forty passengers replaces perhaps thirty cars, and the emissions saved are immediate rather than theoretical. What makes the argument decisive, though, is that road building tends to generate the very traffic it was meant to relieve: new capacity lowers the cost of driving, more people drive, and within a few years the road is as congested as the one it replaced. Investment in public transport does not carry that self-defeating quality.

Three differences, and none of them is vocabulary.

The ideas are extended, not just stated. The first version says pollution is bad and stops. The second explains a mechanism: new roads lower the cost of driving, which generates traffic. Task Response rewards ideas that are developed, and development means one more step of reasoning, not one more sentence.

The cohesion is invisible. The second paragraph contains no Firstly, no Moreover, no Therefore. It holds together through the ideas themselves, through "though" and a colon and a pronoun pointing back. That is what "attracts no attention" means in practice.

The vocabulary is precise rather than impressive. "Self-defeating", "decisive", "capacity" are not rare words. They are exact ones. The band 7 version uses "very beneficial" and "significantly", which are vague words wearing formal clothes. Lexical Resource rewards a word that means exactly what you need, and it penalises a longer word used slightly wrongly far more than it rewards the attempt.

Why so much band 9 advice is wrong

Three habits are taught constantly and all three work against you at the top bands.

1. Memorised introductions. A template opening such as "In this contemporary era, the issue of X has become a topic of heated debate" is recognisable to anyone who reads these essays daily. Memorised language that does not fit the question is explicitly addressed in the descriptors, and it drags Lexical Resource down rather than lifting it.
2. Linking words as a score booster. Candidates are told that cohesive devices raise Coherence and Cohesion. Used mechanically, they do the opposite, because the descriptors ask for cohesion that is not noticeable.
3. Long words for their own sake. Replacing "important" with "paramount" is only an improvement if "paramount" is the right word in that sentence. Usually it is not. Accuracy of collocation, meaning which words naturally go together, is what the top of Lexical Resource is measuring.

How to tell whether an essay you found online is really band 9

Read it with four questions, one per criterion.

1. Does the essay answer every part of the question, including the second half that candidates often miss, and does it hold one position throughout rather than drifting?
2. Can you delete the linking words without the paragraph falling apart? If the argument only holds together because of Firstly and Moreover, cohesion is doing work it should not need to do.
3. Are the unusual words used correctly in their normal collocations, or do they read as inserted?
4. Are there repeated grammar errors, particularly articles, plurals and subject verb agreement? Two of the same error rules out band 9 on that criterion.

Most published band 9 essays fail question 2 immediately.

What to do with this today

Take a body paragraph you have already written. Delete every linking word in it. Read what is left. If the argument still flows, your cohesion is close to the top band. If the paragraph collapses into a list of disconnected sentences, the connections were never in the ideas, and that is the thing to fix before you touch vocabulary.

Then extend one idea by a single step. Whatever your paragraph claims, ask why it is true and write one sentence answering that. That single step is most of the distance between a stated idea and a developed one.

When you want a second opinion on where a paragraph currently sits, you can paste your answer into Bandly and get an estimated band for each of the four criteria with the sentences behind each estimate highlighted. Treat the numbers as estimates from a machine rather than a marked exam paper. Bandly's scoring has not been checked against a qualified human examiner, and on our own sample library the Task Response estimate runs generous, so the criterion worth reading closely is the reasoning underneath rather than the number on top.