How to Describe a Chart or Graph Without Memorizing Phrases
Learn to describe any chart in English by reading the relationship first, choosing accurate trend language, and proving your point with the right data.
Before choosing a graph phrase, decide what the visual is actually doing: changing over time, comparing categories, showing shares, ranking items or staying roughly stable. Then state the main relationship and prove it with one or two numbers.
You understand the chart. Then English asks whether the line surged, soared, climbed, edged up or merely increased—and suddenly the data is easy but the sentence is hard. Reverse the order: read the relationship first; choose the verb second.
Use Relationship First, not phrase first
This is a FunFluen practical teaching method, not an official linguistic or exam taxonomy.
- VISUAL JOB: What is the chart asking you to compare?
- HEADLINE: What is the strongest true relationship?
- PROOF: Which one or two values prove it?
- WORD CHOICE: What is the simplest language that matches that relationship?
That order matters. If you choose the vocabulary first, you can end up forcing the chart into your memorized sentence. If you choose the relationship first, even very plain English can be accurate.
The British Council’s Describing charts guidance follows a compatible principle: introduce what the charts show, give an overview of the key features, then select important patterns or trends rather than describing every visible detail. Relationship First is our way of turning that selection problem into something you can practise repeatedly.
Step 1: identify the visual’s job
Before you write a sentence, ask what kind of relationship the visual contains.
| Visual job | Question to ask | Plain-English starting point |
|---|---|---|
| Change over time | What happened across the period? | It went up / fell / stayed similar / moved up and down. |
| Category comparison | Which groups are higher, lower or similar? | A was higher than B. |
| Part-to-whole | Which categories make up the largest or smallest shares? | A made up almost half of the total. |
| Ranking | What is highest, lowest, second, or close together? | A ranked first, while C was lowest. |
| Mixed pattern | What changed direction or behaved differently? | A rose overall but moved up and down along the way. |
You do not need the chart’s technical name before you can describe its job. If bar chart disappears from your memory, “This visual compares three categories” still gets you moving.
Step 2: say the headline in plain English first
All practice data in this article are fictional. They exist only to teach the language method.
Imagine a line showing monthly users:
| Month | Users |
|---|---|
| January | 40 |
| February | 45 |
| March | 55 |
| April | 70 |
Before searching for a sophisticated trend verb, say what is true:
Plain headline: “The number of users went up every month.”
That sentence is not a beginner failure. It is an accurate analysis. Once the meaning is stable, you can choose a more compact version if you want:
“User numbers rose throughout the period, from 40 in January to 70 in April.”
The second sentence is useful because it compresses the same relationship. It is not better merely because it contains a trend verb.
Step 3: use numbers as proof, not as a transcript
A chart description becomes exhausting when the learner reads every number because they have not decided which numbers matter.
For the fictional sequence 40 → 45 → 55 → 70, you usually do not need four separate sentences. The start and end values already prove the main rise:
“Users increased from 40 in January to 70 in April.”
Add an intermediate value only if it reveals something the headline misses: a slowdown, reversal, sudden jump or unusual gap.
Try the deletion test
Ask of each number: If I remove this value, does the reader still understand the main relationship?
If yes, the value may be detail rather than proof.
This is why an overview matters. British Council guidance recommends selecting key patterns instead of describing everything. A chart is not a hostage situation where every number must be rescued into a sentence.
The Verb Truth Test: make every graph word earn its place
Advanced vocabulary is useful only when the visual supports it. Before using a trend word, ask: What pattern would have to be visible for this word to be true?
Sequence A: 50 → 55 → 60 → 65
Which words pass the truth test?
Supported: rose, increased, rose steadily, increased gradually.
Not supported: fluctuated. The direction never reverses.
Sequence B: 50 → 70 → 55 → 72
Which words pass the truth test?
Supported: fluctuated, moved up and down, ended higher than it started.
Not supported: rose steadily or increased continuously.
Sequence C: 50 → 100
Which words pass the truth test?
Supported: doubled, increased from 50 to 100, increased by 50.
If the baseline and units make a relative comparison useful, this is also a 100% increase. Do not use that wording unless you are sure you mean relative change.
Sequence D: 50 → 80 → 80 → 81
Which words pass the truth test?
Supported: rose and then remained broadly stable; rose before levelling off.
Not supported: kept rising sharply throughout.
Sequence E: 30 → 60 → 45
Which words pass the truth test?
Supported: peaked at 60 during the period, then fell to 45.
“Peak” works because 60 is the highest value in the period shown. Do not use peak merely as a decorative synonym for rise.
A simple verb that is true beats an impressive verb that is beautifully wrong.
Data grammar: tiny words can change the number
Graph English has a few small patterns that carry large differences in meaning.
| What the learner says | Classification | What a listener may understand | Likely intention | Natural alternative | Context note |
|---|---|---|---|---|---|
| “Sales increased from 20 by 30.” | Wrong | The starting value and amount/end point are structurally confused. | Sales moved from 20 to 30. | “Sales increased from 20 to 30.” / “Sales increased by 10.” | From … to … gives start and end; by gives the amount of change. |
| “Sales were increased from 20 to 30.” | Grammatically valid with a different meaning | An agent or action caused sales to be increased. | Neutral observation of a trend. | “Sales increased from 20 to 30.” | The passive can be meaningful when a causative action or agent matters. |
| “The graph is increasing.” | Grammatically valid with a different meaning | The graph itself is becoming larger or increasing. | The value shown on the graph is rising. | “The figure is increasing.” / “The graph shows an increase.” | The literal original can work if the graph itself is physically increasing in size or quantity. |
| “The share increased by 10%” for 40% → 50%. | Context-dependent | A relative 10% increase, which would take 40% to 44%, may be understood. | The share rose ten percentage points. | “The share rose by 10 percentage points, from 40% to 50%.” | From 40% to 50% is a 25% relative increase but a 10-percentage-point increase. The two measures answer different questions. |
Do not let tiny prepositions become decorative. In data language, to and by can change the arithmetic.
Approximate when the visual is approximate
If a bar looks close to 50 but no exact label is printed, saying “exactly 49.7” is not precision. It is fiction with decimal points.
Use bounded language that matches what you can actually read:
- about 50
- roughly half
- just over 30
- just under 100
- nearly twice as high
- around one third
If exact values are printed, use them when they matter. If they are not, do not reverse-engineer confidence from the pixels.
The relationship stays the same; the register changes
Once the analysis is correct, you can adapt the wording for speaking or formal writing without rebuilding your interpretation.
Suppose fictional sales rise from 120 to 190 across the year, with the fastest growth after June.
Spoken presentation
“The main thing to notice is that sales rise across the year, especially after June. They go from about 120 at the start to around 190 by the end.”
Formal written description
“Overall, sales increased across the period, with the strongest growth occurring after June. The figure rose from approximately 120 at the beginning of the year to around 190 at the end.”
The analysis did not change. Only the register did.
For presentations, do not rely on “as you can see” as a substitute for the message. W3C WAI’s accessible presentation guidance advises presenters to communicate relevant visual information aloud and cover the information conveyed by graphics rather than assuming every audience member can access the visual in the same way.
If the chart term disappears, describe the relationship anyway
You do not need to know the exact noun before you can explain the information.
- Forgot bar chart? “This visual compares three categories.”
- Forgot pie chart? “This visual shows how the total is divided among four groups.”
- Forgot peak? “The highest point is 60 in March.”
- Forgot plateau? “After rising, the figure stays at almost the same level.”
- Forgot fluctuate? “It goes up and down during the period.”
That is not cheating. The technical label is useful shorthand; the relationship is the actual message.
Practice: the Cold Chart Lab
Every dataset below is fictional practice data. Do not search for a phrase first. Before opening the model, answer four questions:
- What is the visual job?
- What is the headline relationship?
- Which one or two values prove it?
- What is the simplest accurate sentence?
Cold Chart A: change over time
| Month | Users |
|---|---|
| January | 40 |
| February | 45 |
| March | 55 |
| April | 70 |
Compare a possible analysis
Visual job: change over time.
Headline: users increased throughout the period.
Proof: 40 in January → 70 in April. The middle values confirm the rise but do not all need separate sentences.
Plain model: “Monthly active users increased from 40 in January to 70 in April.”
More compact trend wording: “Monthly active users rose throughout the four months, from 40 to 70.”
Cold Chart B: category comparison
| Plan | Users |
|---|---|
| Basic | 62 |
| Plus | 38 |
| Pro | 21 |
Compare a possible analysis
Visual job: category comparison/ranking.
Headline: Basic has the most users, while Pro has the fewest.
Proof: 62 versus 21 is enough to prove the extremes; add Plus only if the middle position matters.
Model: “Basic was the most widely used plan at 62 users, compared with 38 for Plus and 21 for Pro.”
Cold Chart C: part-to-whole
| Source | Share |
|---|---|
| Search | 48% |
| Social | 31% |
| Direct | 21% |
Compare a possible analysis
Visual job: part-to-whole comparison.
Headline: Search contributes the largest share, at almost half of the total.
Proof: 48%; comparison with 31% shows a clear lead.
Model: “Search accounted for 48% of traffic, making it the largest source, ahead of Social at 31%.”
Cold Chart D: movement with reversals
| Quarter | Index |
|---|---|
| Q1 | 50 |
| Q2 | 70 |
| Q3 | 55 |
| Q4 | 72 |
Compare a possible analysis
Visual job: change over time with fluctuation.
Headline: the index fluctuated but ended higher than it began.
Proof: rise to 70, fall to 55, recovery to 72.
Model: “The index fluctuated across the year, rising from 50 to 70, falling to 55, and then recovering to 72.”
“Rose steadily” fails the Verb Truth Test because the direction reverses.
Cold Chart E: rise then stability
| Week | Score |
|---|---|
| 1 | 50 |
| 2 | 80 |
| 3 | 80 |
| 4 | 81 |
Compare a possible analysis
Visual job: change over time.
Headline: the score rose sharply at first and then remained broadly stable.
Proof: 50 → 80, then 80 → 81 over the remaining weeks.
Model: “The score rose from 50 to 80 in the second week, then levelled off at around 80.”
Micro-challenge: One Chart, Three Voices
Take one chart you already have. Keep the underlying claim fixed and describe it three ways:
- Plain English: one sentence with no specialist trend vocabulary.
- Spoken: a 10-second version for someone listening to you present.
- Formal written: two sentences with an overview and one proof point.
If the core meaning changes between versions, the problem is not register. Re-read the relationship.
If you are preparing for IELTS, this is the underlying skill—not the whole Task 1 strategy
Current official IELTS Academic Writing guidance says Task 1 may ask you to describe visual information such as graphs, charts, tables or diagrams in your own words. The task requires at least 150 words, candidates are advised to spend about 20 minutes, and the style should be academic or semi-formal/neutral.
This article teaches the language-generation engine underneath that task: select the important relationship, support it accurately and choose language that matches the data. It does not teach band-score tactics, a Task 1 template or exam-specific timing strategy.
For spoken practice, make the chart unfamiliar again
Once you can do the Cold Chart Lab in writing, speaking adds a different pressure: you have less time to search for the “perfect” phrase.
You can practise saying the main relationship before reaching for a polished phrase by choosing a speaking-practice path in FunFluen. Bring one of the fictional datasets above or one of your own charts: state the headline relationship aloud, then add one proof point.
The destination provides general English speaking practice. This exact chart task is not preloaded, FunFluen does not inspect or analyse the chart for you, and it does not guarantee fluency.
The chart can change. Your method should survive.
A memorized sentence feels powerful until the next visual has no dramatic rise, no convenient peak and no place for the phrase you rehearsed.
A relationship-first description is harder to break. Ask what the visual is doing. State the strongest true relationship. Prove it with the few numbers that matter. Then choose the simplest language that passes the truth test.
Meaning first. Vocabulary second. The chart gets the final vote.
Sources
Explore more language-learning guides in Media-Based Language Learning.