Writing more essays is not the same as improving
Practice only works when each attempt corrects the last one. If you write essay after essay and no one points out what went wrong, you aren't building skill — you're building habit. The same dropped articles, the same vague position, the same missing half of the question appear in essay one and essay thirty. You get faster at producing the exact writing that capped you at 6.5.
Volume feels like effort, so it's reassuring. But the examiner doesn't reward effort. They mark what's on the page against four fixed descriptors, every time. Twenty unmarked essays move you nowhere; one essay you genuinely understand can move you a lot.
The broken feedback loop
Here's the loop most candidates are stuck in: write an essay, sense it wasn't great, write another one hoping it's better. Nothing in that cycle tells you what was weak or why. You can't fix an error you can't see — and your own brain is the worst place to spot it, because the mistake looked correct when you wrote it.
Real improvement needs a loop that closes: a specific error named, a specific fix made, and proof the fix worked. Miss any one of those and you're guessing. Most people are missing all three.
Chasing fancy words while the real problems hide
The most common wrong turn is reaching for impressive vocabulary. Candidates memorise "plethora", "myriad" and "detrimental" and sprinkle them in, convinced bigger words mean a bigger band. They don't. A less-common word used slightly wrong counts as an error, not a flourish — and it can drag Lexical Resource down rather than up.
Meanwhile the two criteria quietly capping you go unaddressed. Task Response: you didn't fully answer the question, or your position stayed fuzzy. Grammatical Range and Accuracy: small slips repeat in every paragraph — a missing article here, a wrong preposition there, a subject that doesn't agree with its verb. Fancy vocabulary can't rescue an essay that didn't answer the prompt or that leaks the same grammar error a dozen times.
Why so many people plateau at exactly 6.5
The 6.5 plateau is so common it's almost a signature. It usually means your ideas are decent and your essays read fluently — but two things keep recurring. You answer most of the question instead of all of it, so Task Response stalls. And accurate, error-free sentences are outnumbered by sentences with one small slip, so Grammatical Accuracy stalls too. Average four criteria where two are stuck and you land on 6.5 again and again.
The frustrating part: from the inside, a 6.5 essay feels like a 7 essay. It's coherent, it's on-topic, it sounds confident. The gap is in details you've stopped noticing — which is exactly why writing more of them changes nothing.
What actually moves the band
Improvement isn't more essays. It's a tighter loop on the essays you already have:
1. Identify your single most frequent error pattern. Not ten things — the one that shows up most. Maybe it's dropped articles ("government should fund the arts"). Maybe it's a position you never commit to. Maybe it's only answering one of the two parts the question asked.
2. Make the smallest targeted fix. Don't rewrite the whole essay. Correct that one pattern wherever it appears. Add the missing articles. Sharpen the position into one clear sentence. Add the paragraph that answers the part you skipped.
3. Re-mark the same essay. Run the corrected version through the same strict marking and watch whether the band actually moves. This is the step almost everyone skips — and it's the only one that proves your fix was real instead of imagined.
4. Repeat. One pattern at a time, one re-mark at a time. This is how a stalled band starts climbing.
Honest measurement beats a flattering score
None of this works if the feedback lies to you. Most free checkers and general AI tools inflate the band, wave through whether you answered the question, and hand you a vague number that feels good and teaches nothing. A flattering score is worse than no score — it tells you to keep doing exactly what's keeping you stuck.