The official band descriptors tell you the criteria. This page translates those criteria into what they actually look like on the page — the practical, visible differences between a Band 7, a Band 8, and a Band 9 Task 1 response. All examples below are written from scratch to illustrate the point; they are not from any real exam script.
The One Idea That Matters Most
Higher bands are not "more advanced" writing. They are more controlled writing.
A candidate who writes plain, accurate sentences about the data will consistently outscore a candidate who reaches for impressive vocabulary and gets it wrong. Task 1 rewards a writer who says exactly what they mean, every time, over a writer who says something more ambitious some of the time.
This shows up in three places: the overview, the data selection, and the sentence structures. A response can be weak in one of these and still hit Band 6–7. It cannot be weak in any of them and reach Band 8–9.
1. The Overview
The overview is the single biggest gap between Band 6 and Band 7+. It is also the first thing an examiner checks.
What a missing or weak overview looks like: the response jumps straight into describing numbers in the order they appear on the chart, with no sentence anywhere that steps back and says what the data means as a whole.
What a Band 7 overview looks like: a clearly separate sentence or short paragraph, placed early (usually right after the paraphrase, before the detailed data), that names the 2–3 biggest patterns — without citing specific figures. For example, for an imaginary chart on commuting methods in three cities:
Overall, cycling was the most common way to commute in City A, while car use dominated in City B and C, and public transport played only a minor role everywhere.
That sentence is correct and sufficient for a 7 — it picks out real trends and stays selective.
What pushes it to Band 8–9: the overview is not just present but precise — it captures the pattern that actually organizes the data (a comparison, a turning point, a split between two groups) rather than a generic list of "some went up, some went down." A Band 9 overview reads as if the writer understood the chart's whole shape before writing a single number.
2. Data Selection and Grouping
Weaker responses try to mention every number on the chart, in the order given, with no grouping. This produces a list, not a description.
Band 7 responses group related data together — by category, by time period, or by whichever axis makes the comparison clearest — rather than following the chart left to right mechanically. They also make a deliberate choice about which figures to use as supporting evidence and which to leave out, since not every data point needs to be named.
Band 8–9 responses go further: the grouping itself demonstrates understanding of the data's structure. A strong response might organize a four-line graph into "the two lines that rose" and "the two that fell," or a table into "categories that changed a lot" and "categories that stayed flat" — a structure that only becomes obvious once you've actually understood the relationships in the numbers, not just read them off.
For process diagrams and maps, the equivalent skill is sequencing: Band 7+ responses use clear stage markers (first, following this, at the final stage) and group sub-steps under the main stage they belong to, rather than describing every arrow as a separate, disconnected event.
3. Sentence Variety — and the Trade-Off With Accuracy
This is where candidates most often damage their own score by overreaching.
- Band 6 typically shows a mix of simple and complex sentences, but complex ones are noticeably less accurate than the simple ones — the writer's control drops as soon as the sentence gets longer.
- Band 7 shows a real variety of structures (comparatives, relative clauses, linking devices, passive constructions where natural) used with reasonable accuracy, and — this is the key part — a few errors that never obscure meaning.
- Band 8–9 shows that same variety used so consistently accurately that the sentence structure itself stops being noticeable. The reader focuses on the data, not on the grammar.
A useful test: if a sentence would be more accurate in a simpler form, and the complex form adds no extra precision, the complex form is not helping the score. Compare:
Weak attempt at complexity: "The figure which represents for the sales in Region A, that had been fluctuating, then it increased dramatically at the end of the period."
Same idea, controlled: "Sales in Region A fluctuated for most of the period before rising sharply at the end."
The second version is shorter, contains one clear comparative structure, and has zero grammar errors. That is what examiners mean by accuracy beating complexity — it is not an argument for writing only simple sentences, it is an argument for not attempting a structure you cannot yet control under exam pressure.
4. Task-Specific Signals
Line graphs, bar charts, tables (Academic): the overview names trends, not numbers. Detail paragraphs use comparison and contrast language (whereas, in contrast, similarly) to link related figures rather than presenting them as isolated facts.
Process diagrams and maps (Academic): the overview names the type and scale of the process (how many stages, whether it's cyclical or linear; for maps, the overall nature of the change). Sequencing language replaces trend language.
Letters (General Training): there is no "overview" in the numeric sense — its equivalent is a clear, consistent purpose and tone established in the opening lines and held throughout. A Band 7+ letter doesn't drift from formal to casual (or the reverse) partway through, and it addresses every bullet point without treating any of them as an afterthought.
Quick Self-Check
| Question | If "no," you are likely capped at |
|---|---|
| Is there one sentence (or short paragraph) that summarizes the big picture, with no numbers in it? | Band 6 |
| Does it appear early, not at the end? | Band 6–7 |
| Are the data points grouped logically rather than listed in chart order? | Band 6–7 |
| Do the complex sentences you attempted stay accurate, not just ambitious? | Band 7 |
| Could a reader follow every sentence on the first read, without re-parsing it? | Band 7–8 |
| Does the response address the exact task given, with nothing generic bolted on? | Band 8–9 |
If you can answer "yes" to all six honestly, the remaining gap to a 9 is consistency across a full response, not a missing technique — which is a matter of timed practice, not new information.