How to get IELTS writing feedback when you don't have a teacher

If you are practising IELTS writing without a teacher, the most useful feedback you can get is not encouragement — it is someone, or something, showing you the faults you cannot see in your own essays. Without that, solo practice tends to reproduce the same problems essay after essay, which is one of the most common reasons candidates stay stuck around band 6.5.

This guide covers why your own essays are hard to judge, four realistic ways to get feedback without a teacher, what each one can and cannot tell you, and a simple test for whether a piece of feedback is worth acting on.

Feedback is useful when it finds the line you could not see.

How to get IELTS writing feedback without a teacher: start with what you can't see

Writing more essays on your own does help with speed and stamina. It does much less for your band, because an essay you cannot evaluate teaches you very little. You finish it, it feels reasonable, and you write the next one the same way — with the same thin paragraph, the same repeated linking word, the same article error in the same kind of noun phrase.

So the goal of feedback is narrow and practical: to find the specific features that are holding each of the four criteria down. An examiner marks every Task 2 essay on Task Response, Coherence and Cohesion, Lexical Resource, and Grammatical Range and Accuracy (in Task 1, the first criterion is Task Achievement). Each criterion gets a whole band, and the four are averaged into the task band. That means one weak criterion can pull the whole task down even when the other three are solid.

Feedback that helps you tells you which criterion that is, and why. Feedback that does not — however kind or detailed it sounds — leaves you where you started. For a closer look at how that pattern shows up around 6.5, see why am I stuck at band 6.5 in IELTS writing.

Why you can't see the problems in your own essay

This is not a question of effort or intelligence. It is how reading your own writing works.

You read your intention, not the page. When you reread a paragraph, you already know what it was supposed to prove. Your mind fills in the step you skipped, the example you meant to connect, the part of the question you planned to address. An examiner only has the page. This is why candidates so often overrate their Task Response: the essay answers the question in the writer's head, but only partly on paper.

Familiar errors look correct. If you have written "the most of people" or "it can cause to increase" many times, the phrase stops looking unusual to you. Habitual errors are, almost by definition, the ones you do not notice. They are also the ones that repeat, and repetition is what an examiner reads as a pattern in Grammatical Range and Accuracy.

Your own linking feels natural. Starting three paragraphs with "Firstly", "Secondly" and "Moreover" feels organised when you write it. A reader may experience it as mechanical, and Coherence and Cohesion is marked from the reader's side.

You compare against your last essay, not the descriptor. Solo practice tends to measure progress relative to yourself: this one felt easier, this one used better words. The band is measured against fixed descriptors. An essay can be your best so far and still sit at the same band as the last five.

None of these can be fixed by rereading more carefully. They need a second reader — human or automated — who does not share your intentions and habits.

Four feedback routes, and what each one misses

Each route answers a different question about the same essay.

There is no single route that does everything a good teacher does. The practical approach is to know what each one is good at and combine them.

1. A study partner or essay exchange

Swapping essays with another candidate is free and genuinely useful. A partner is a real reader who does not know what you meant to say, so they are well placed to tell you:

• where they lost the thread of your argument
• which sentence they had to read twice
• what they think your position is, in one sentence

If their summary of your position is not what you intended, you have found a Task Response or Coherence problem without anyone needing to know the descriptors.

The limit is level. A partner at a similar band tends to share many of your blind spots, so they are unlikely to catch errors they make themselves, and they cannot tell you with confidence whether an essay has moved from 6.5 to 7. Use partner feedback for clarity and argument, not for a band.

2. A general AI assistant

A general assistant is quick and available at any hour, and it is good at narrow, checkable tasks: listing grammar errors in a table, suggesting alternatives for a repeated word, or restating your main argument.

It is weaker at the part that matters most for many candidates at 6.5. Asked for "a band", a general assistant tends to respond to how fluent the essay sounds and does not reliably test whether each part of the question has been fully addressed. That is why its scores often come out higher than an examiner's, usually on Task Response. We look at this in detail in does ChatGPT overrate your IELTS writing.

3. A descriptor-based checker

A checker built around the public band descriptors — this site is one — marks each of the four criteria separately and ties its judgements to specific sentences. That gives you what the first two routes struggle to: a criterion-level estimate you can use to decide what to work on next, available every time you write.

It has limits too, and they are worth stating plainly. The result is an estimate against the public descriptors, not an official band. It cannot talk through your reasoning the way a person can, and it is only as useful as the essay you give it: a piece written slowly with a dictionary open will be scored on that piece, not on your test-day writing. For how automated scores compare with examiners, see how accurate are AI IELTS writing scores.

4. A paid tutor for occasional sessions

Even without a regular teacher, a tutor who knows the descriptors can be worth an occasional session. A person can ask why you structured a paragraph a certain way, notice a habit across several essays, and explain a descriptor line in terms of your writing.

The limits are cost and frequency. Most self-study candidates cannot get tutor feedback on every essay, and quality varies — a tutor who comments mainly on style or "sophisticated vocabulary" gives you the same weak feedback as anyone else. If you use a tutor, bring them a few essays that have already been through the other routes, and ask them about the criterion that keeps coming out lowest.

A reasonable combination for most self-study candidates: write under timed conditions, get a criterion-level estimate on every essay, swap one essay a week with a partner for a reader's view, and save tutor time, if you have it, for the criterion that will not move.

What useful feedback actually looks like

Whatever the route, you can test a piece of feedback in seconds. Useful feedback does three things: it names a criterion, it points to the sentence that caused the problem, and it says what to do next time. Feedback that does none of these may be true, but it is very hard to act on.

Here are three comments of the kind candidates often receive, and why they do not help.

• "Your essay sounds a bit awkward in places." No criterion, no sentence, no action. Awkward could mean word choice, grammar, or linking, and you cannot tell which.
• "Good vocabulary, try to use more advanced words." This pushes you towards rare words used imprecisely, which can hurt Lexical Resource rather than help it.
• "Overall this is around a 6.5 or 7." A single number without reasons tells you nothing about which criterion to work on.

And three versions of the same observations that you can act on.

• Coherence and Cohesion: "Paragraphs two, three and four each open with a linking word ('Firstly', 'Moreover', 'In addition') before any idea is stated. Next time, open each body paragraph with its main point and let the link come from the content."
• Lexical Resource: "'Detrimental repercussions' in your second paragraph does not fit the everyday example that follows. 'Serious effects' would be more precise. Choose the word that matches the idea, not the most unusual one."
• Task Response: "The question asks whether the advantages outweigh the disadvantages. Your essay lists both but never says which is greater until the last sentence of the conclusion. State your answer in the introduction and support it in every body paragraph."

Notice that the useful versions are not more critical. They are simply more specific. Once you start sorting feedback this way, you will also find it easier to give better feedback to a study partner, which makes the exchange more valuable for both of you.

Get criterion-level feedback on your next essay

The simplest way to put this into practice is to write one essay the way you would on test day — an unseen question, 40 minutes, no editing afterwards — and then get feedback that meets the three tests above: a band per criterion, the sentences behind each judgement, and a clear starting point.

When the result comes back, start with the lowest criterion, not the overall estimate. Look at the sentences it points to and ask whether the same feature appears in your last few essays. If it does, that is the habit to work on for the next week. Then write the next essay, check it the same way, and see whether that criterion has moved.

Remember that this is an estimate against the public descriptors, not an official band. Its value is in showing you, on every essay, the part of your writing you could not see on your own. Once you know which criterion is weakest, the next question is how to prioritise between them; our upcoming guide on which criterion to fix first covers that. For what each criterion needs at the next level, see how to get band 7 in IELTS Writing.

FAQ

How can I get IELTS writing feedback without a teacher?

Combine routes rather than relying on one: swap essays with a study partner for a reader's view, use a checker that marks against the public band descriptors for criterion-level estimates, and check your own essay against a short list tied to the four criteria. Judge every comment by whether it names a criterion, points at a sentence and tells you what to change.

Why can't I see the mistakes in my own IELTS essay?

Because you read your own intention, not the text on the page. You know what each paragraph was meant to prove, so gaps in Task Response are invisible to you, and errors you make habitually look correct because you have seen them many times. This is how reading works, not a lack of effort.

Is peer feedback useful for IELTS writing?

Yes, for some things. A partner can tell you where they lost the thread, which sentence they had to reread, and what they think your position is. A partner at a similar level is less reliable at spotting the errors you share, and cannot tell you with confidence whether an essay has reached the next band.

Can ChatGPT give me IELTS writing feedback?

It can help with specific, checkable questions, such as listing grammar errors or restating your position in one sentence. It is less reliable when asked for a band, because a general assistant tends to judge how fluently an essay reads rather than whether it fully answers the question.

What does good IELTS writing feedback look like?

It names one of the four criteria, quotes or points to the sentence that caused the problem, and says what to do differently in the next essay. Comments such as "sounds awkward" or "good vocabulary" do none of those things, so they rarely change the next score.

Related reading

• Why am I stuck at band 6.5 in IELTS writing? — four real causes, criterion by criterion
• Does ChatGPT overrate your IELTS writing? — where a chat score and an examiner part ways
• My essay looks band 7 but scored 6 — why your own reading of an essay runs high
• IELTS Writing band descriptors explained — the full wording for all four criteria

The four criteria named here are those in the public Writing band descriptors, published by IELTS and the British Council.