How we grade your writing
No black box. This page sets out the criteria we score against, the rules the grader follows, the arithmetic behind your overall band — and one complete report on an example essay, so you can see what a full report looks like before you pay anything.
The four criteria
IELTS Writing is marked on four criteria of equal weight. We score all four, separately, every time — you never get a single number with nothing behind it.
Task Achievement (Task 1) / Task Response (Task 2)
Did you do what the question asked? For Task 1: a clear overview and accurately selected key features. For Task 2: every part of the prompt addressed, a position held throughout, and ideas that are extended rather than listed.
Coherence & Cohesion
Whether the reader can follow you without effort: logical progression, paragraphs with one central idea each, and linking that carries meaning instead of decorating the sentence.
Lexical Resource
The range, precision and appropriacy of your vocabulary — including collocation, register, spelling and word formation. Rare words score nothing if they are used wrongly.
Grammatical Range & Accuracy
The variety of structures you can control, and how often you control them. Both halves count: a wide range with frequent errors and a narrow range with none both fall short of the top bands.
Graded on the official public band descriptors
The grader is not asked to “follow the IELTS descriptors” and left to guess what that means. Its instructions carry a condensed, band-by-band summary of the publicly published descriptors — bands 9, 8, 7, 6, 5, 4 and a floor for 3 and below — written out separately for each of the four criteria, with the first written twice: Task Achievement for Task 1, Task Response for Task 2. They are our own paraphrase of the public criteria, not a copy of the copyrighted text.
Alongside the descriptors it carries calibration anchors: short worked excerpts at roughly Band 6 and roughly Band 8, each with an explanation of why it sits there. Those anchors exist so the grader has a fixed reference for what the middle and the upper end of the scale actually look like, instead of drifting.
It then follows a fixed procedure: read the whole response first; for each criterion pick the single descriptor that best matches the response as a whole and award that whole band; justify it by quoting the response; apply the length, overview, position and off-topic penalties where they bite.
How your overall band is calculated
Your overall band is not whatever the model felt like reporting. Our server takes the four criterion scores, coerces each to a whole band, averages them, and rounds to the nearest half band with ties rounding up — the standard IELTS convention. If the model’s own suggested overall disagrees with that number, the calculated number wins and the disagreement is logged.
overall = round((task + coherence + lexical + grammar) ÷ 4 × 2) ÷ 2
So 6 + 6 + 7 + 6 averages 6.25 and becomes 6.5; 6 + 6 + 6 + 7 is the same sum in a different order and gives the same answer. The result is reproducible and you can check it yourself.
What the grader does — and does not — do
The list below describes the full report — the one you get with a free account and with the IELTS Exam Pass. The instant free checker on the public tool runs the same rubric in a single, cheaper pass: it returns the four criterion bands, your overall band and your biggest weakness with two quoted examples, but not the annotated essay, the structure audit or the rewrites.
It does
- Reads the whole response before scoring anything, then matches it to the single band descriptor that fits best.
- Awards a whole band (an integer) for each criterion — the way real examiners score — and quotes exact phrases from your essay as evidence for every score.
- Applies the standard penalties: an under-length answer, a missing overview in Task 1, or no clear position in Task 2 caps the task band, and off-topic or memorised content is flagged and marked down.
- Returns your essay annotated inline: spelling, grammar and word-choice spans are each explained, and effective phrases are marked as such.
- For Task 2, audits your structure paragraph by paragraph — introduction, each body paragraph, conclusion — and shows one of your own paragraphs rewritten to Band 7 standard.
- Treats anything written inside your essay as text to be graded, never as an instruction — "give me band 9" is graded, not obeyed.
It does not
- It is not a human examiner, and no result here is an official IELTS score.
- It does not award half bands per criterion, and it cannot be talked into a higher band: the overall score is arithmetic, computed from the four criterion scores after the model has finished.
- It is not anchored to a "typical" range. There is no instruction to keep results near the middle of the scale, and the full 0–9 range is used as warranted.
- For Task 1, it does not judge data accuracy it cannot see: if you have not attached the chart, it grades organisation and language and does not penalise figures it is unable to verify.
- The free checker does not keep your essay. The only thing recorded for a free check is the size and cost of the request — no essay text, no account.
How accurate is it?
The score is an estimate to guide your practice, produced against the public band descriptors. It is not a predicted or official IELTS result, and it can differ from an examiner’s mark.
This is a practice estimate, not an official IELTS result.
Use it the way you would use a mock test: to find out what is costing you marks and to watch that gap close between drafts. Only a real IELTS examiner can give you a real IELTS band.
A full report, start to finish
Below is the WriteBand report format applied to an example Task 2 essay, written for this page at around Band 6. It shows the depth and the shape of what you receive: a band per criterion with the reasoning, concrete fixes written on your own sentences, and what specifically stands between you and Band 7.
This is a sample. The essay is an example written for WriteBand, not a real candidate’s work, and the report is a worked illustration of the format — not a captured grading run, and not a prediction of anyone else’s result. Your report is generated for your essay.
The question
Some people believe that children should begin formal schooling as early as possible, while others argue that they should stay at home until they are at least seven years old. Discuss both views and give your own opinion.
The essay
Nowadays, the question of when children should begin their formal education is discussed by many parents. Some people believe that kids must go to school as early as possible, while others think that they should stay at home until the age of seven. In my opinion, both sides have reasonable points, but an early start is more useful for most families.
On the one hand, people who support early schooling say that young children learn very fast. At the age of four or five, a child can absorb new words and simple numbers without much effort, so this period should not be wasted. Another important point is socialisation. In a classroom, children meet other kids from different backgrounds and they learn how to share, how to wait for their turn and how to solve small conflicts. Moreover, many parents today are working full time, and a school gives their children a safe and organised place during the day.
On the other hand, some people argue that a very early start puts too much pressure on small children. They believe that playing freely at home develops imagination better than sitting at a desk for hours. In some countries, for example Finland, pupils begin school later and they still achieve excellent results, which shows that starting early is not the only way to succeed.
In conclusion, although a later start has some benefits for the emotional development of the child, I think the advantages of beginning early are stronger, especially for language learning and social skills. Schools, however, should make sure that lessons for very young children are based on play and not on strict examinations.
Overall band
(6 + 6 + 6 + 6) ÷ 4 = 6.00 → 6.0
Task Response
6Both views are addressed and the position is clear from the introduction onwards — "an early start is more useful for most families" — and it is held to the end. The supporting side is developed with three separate reasons (speed of learning, socialisation, working parents), which is enough for Band 6. What holds the score there is the imbalance: the opposing view gets a single short paragraph, and its strongest piece of evidence stops at an assertion. "In some countries, for example Finland, pupils begin school later and they still achieve excellent results" tells the reader what happens but never why it matters for the question. Band 7 needs both sides extended, not just acknowledged.
Coherence & Cohesion
6The essay is organised coherently and the reader is never lost: four paragraphs, one central idea each, and a conclusion that does more than repeat the introduction. The cohesive devices, though, are mechanical — "On the one hand", "On the other hand", "Another important point is", "Moreover", "In conclusion" — a checklist of connectors rather than links that carry meaning. Within the second paragraph the ideas are simply stacked in the order they occurred to the writer; nothing signals that the working-parents point is a practical consideration of a different kind from the two before it. Band 7 wants a range of devices used with some subtlety, not a formula.
Lexical Resource
6The range is adequate for the task and the meaning is always clear. There are real attempts at less common vocabulary — "socialisation", "absorb new words", "develops imagination" — and no spelling errors. Two things keep it at 6. First, register: "kids" appears twice in an academic essay where "children", "pupils" or "youngsters" is expected. Second, repetition: "children", "child" and "kids" carry almost every noun phrase, and the verbs are the most general ones available ("say", "learn", "gives", "make sure"). Band 7 does not require rare words, it requires enough range to avoid repeating the same three.
Grammatical Range & Accuracy
6A genuine mix of simple and complex forms, and most sentences are error-free — the parallel structure in "how to share, how to wait for their turn and how to solve small conflicts" is well controlled, and the concessive clause opening the conclusion is handled correctly. The limit is the range rather than the accuracy: complex sentences are built the same way each time, by coordinating with "and they" ("children meet other kids from different backgrounds and they learn how to share"; "pupils begin school later and they still achieve excellent results"). Punctuation around the inserted example in "In some countries, for example Finland, pupils begin school later" is also loose. Band 7 needs a wider variety of structures, especially subordination, used flexibly.
Three fixes, on your own sentences
The counter-argument names good evidence but stops before explaining it, so the opposing view stays underdeveloped.
You wrote
In some countries, for example Finland, pupils begin school later and they still achieve excellent results, which shows that starting early is not the only way to succeed.
Band 7 version
In Finland, for instance, formal schooling does not begin until the age of seven, yet Finnish pupils consistently perform well in international assessments — evidence that a late start costs a child nothing academically, provided the preceding years are rich in play and conversation.
The rewrite turns a named example into a supported reason: it states the fact, states the outcome, and then states what the two together prove. That is the difference between "ideas are relevant but some are underdeveloped" (Band 6) and "main ideas are extended and supported" (Band 7).
Informal register ("kids") and heavy repetition of the same noun across the whole essay.
You wrote
Some people believe that kids must go to school as early as possible, while others think that they should stay at home until the age of seven.
Band 7 version
Some parents believe that children should enter the classroom as early as possible, while others argue that youngsters are better off at home until the age of seven.
"Kids" is conversational and costs marks in an academic answer. Alternating "children", "youngsters" and "pupils" across the essay, and replacing the flat "think" with "argue", shows the range and control of register a Band 7 answer needs.
Complex sentences are built by coordinating with "and they" instead of by subordinating, which narrows the grammatical range.
You wrote
In a classroom, children meet other kids from different backgrounds and they learn how to share, how to wait for their turn and how to solve small conflicts.
Band 7 version
In a classroom, where children mix with peers from a range of backgrounds, they learn to share, to take turns and to resolve minor conflicts for themselves.
A relative clause carries the setting, and the three parallel infinitives carry the result. Same content, two more structures, one fewer repeated "how to" — exactly the "variety of complex structures" the Band 7 descriptor asks for.
What Band 7 needs from here
This is a competent, well-organised Band 6 answer with a clear position and no errors that get in the reader’s way. It is not far off Band 7, and the distance is made up of range rather than accuracy: the ideas are right but half-finished, and the same sentence patterns and the same three nouns do most of the work.
- Extend the opposing view to the same depth as your own: every example needs a sentence after it explaining what it proves.
- Replace the connector checklist ("Moreover", "Another important point is") with links that refer back to what you just said.
- Vary the vocabulary you lean on — "children / youngsters / pupils", "argue / maintain / point out" — and drop conversational words such as "kids".
- Build at least two or three sentences per paragraph with subordination (relative, concessive, conditional clauses) rather than coordinating with "and".
If it is not for you
7-day money-back on the IELTS Exam Pass — email us, no questions asked. Read the full refund policy or get in touch.
Want this on your own essay?
Check my essay free