PTE Describe Image: Pie Charts, Process Diagrams & Maps — Advanced 79+ Strategies (2026)

If you score 65–74 in PTE Speaking and Describe Image (DI) feels inconsistent, this post is specifically for you. You likely handle bar charts and line graphs with a practised template — but the moment a pie chart, a process diagram, or a map appears, your 40-second window disappears before you've said anything meaningful. Pie charts, process diagrams, and maps together account for roughly 35–40% of all DI tasks in live PTE tests in 2026. Leaving them under-prepared is leaving points on the table.

Our existing guide — PTE Describe Image: Advanced Data Narration for Multi-Trend Line Graphs & Comparative Bar Charts (2026) — covers the two most complex data chart types. This guide fills the remaining gap: the three non-data-chart image types that require entirely different cognitive and linguistic frameworks.

Why Pie Charts, Process Diagrams & Maps Fail Standard Templates

The classic DI template — "The image shows... The most significant trend is... Overall..." — was designed for time-series charts where you move through data chronologically. It breaks down for three reasons depending on image type:

Pie Charts: There is no timeline. All segments are simultaneous and proportional. Students using a trend-analysis template try to describe each slice sequentially without making comparisons — the very skill the AI rewards.

Process Diagrams: There is no data to compare. Everything is sequential and causal. Students applying a data-description template say "the image shows..." and immediately stall because there are no numbers, no axes, and no trends.

Maps: There are multiple spatial relationships operating simultaneously (location, direction, proportion, change). Students without a spatial-language framework describe individual features in isolation, missing the geographic logic that earns Content 4.5–5.0.

The PTE AI Content rubric rewards relationship communication — showing you understand how the elements of the image connect. A different image type requires a different relationship framework.

Framework 1: SLICE — PTE Describe Image Pie Charts

Pie charts appear in PTE DI in two main formats: a single pie chart showing proportional breakdown of one dataset, and a comparative pair (two pie charts, same categories, different years or groups). The SLICE framework addresses both.

S — Summarise the whole: State what the pie chart represents in one sentence. Name the total (if given) or the category being divided.

L — Largest slice: Identify and name the dominant segment first. State its percentage. This is your anchor.

I — Interesting comparison: Name the second or third significant segment. If it's a double pie, name the most changed segment across the two charts.

C — Contrast the minority: Name the smallest or most surprising segment. This shows evaluative depth.

E — Express the overall pattern: Close with a one-sentence synthesis that describes the distribution logic — is it concentrated (one dominant segment >50%) or distributed (several roughly equal segments)?

SLICE Applied: Single Pie Chart

Hypothetical image: Pie chart titled "Global Freshwater Usage by Sector." Segments: Agriculture 69%, Industry 19%, Domestic 12%.

SLICE response:

"The pie chart illustrates the global distribution of freshwater consumption across three sectors. Agriculture is by far the dominant user, accounting for 69% — more than two-thirds of total usage. Industry represents a secondary but significant share at 19%, while domestic consumption accounts for the smallest proportion at just 12%. Overall, freshwater consumption is heavily concentrated in the agricultural sector, with the remaining 31% split between industrial and household use."

Word count: ~85 words | Time: ~35–36 seconds | Content score: 4.5–5.0

SLICE Applied: Double Pie Chart

Hypothetical image: Two pie charts — "Energy Mix 2000" and "Energy Mix 2020." Categories: Coal, Gas, Renewables, Nuclear, Oil.

Key adaptation for double pie: The L and I steps compare change, not just size.

"The charts compare the composition of global energy generation in 2000 and 2020. In 2000, coal was the dominant source at 38%, followed by oil at 26%. By 2020, coal had declined to 27% — a reduction of 11 percentage points — while renewables had risen from 8% to 21%, the most dramatic proportional shift across the period. The overall pattern shows a measurable transition away from fossil fuels, though coal and oil together still represented more than half of the 2020 energy mix."

What changed: L and I anchor on the biggest shift rather than the current biggest slice. This is what Content 4.5–5.0 responses do differently.

Common Pie Chart Mistakes

Mistake 1: Listing every segment. A pie chart with six segments cannot be fully enumerated in 40 seconds without sacrificing the comparative language the AI rewards. Target: name 3 segments maximum unless the chart has only 3–4 segments total.

Mistake 2: No proportional language. Saying "Agriculture is 69%" is weaker than "Agriculture accounts for almost 70% — more than two-thirds of total consumption." Fractions and comparative expressions signal Content depth.

Mistake 3: Missing the overall sentence. Students who run out of time often cut the E step. Don't. The closing synthesis is worth more to the Content rubric than naming one additional segment. If time pressure forces a choice, drop one middle segment rather than the closing sentence.

Framework 2: STAGE — PTE Describe Image Process Diagrams

Process diagrams (also called flowcharts or cycle diagrams) appear in two main variants: linear processes (a sequence with a clear start and end) and cyclical processes (a continuous loop with no fixed start). The STAGE framework applies to both.

S — State the subject: What process is being described? Name it explicitly and completely in one sentence.

T — Total steps: State how many stages the diagram shows. This one sentence gives the AI structural context and buys you credibility for the rest of the response.

A — Anchor the starting point: Name the first stage clearly. For cycles, choose the stage most logically associated with initiation.

G — Go through the key sequence: Cover 3–4 stages using causal/sequential connectors: "which then leads to," "this results in," "following this," "the subsequent stage involves." You do NOT need to name every stage — cover the ones that show the logical progression.

E — Explain the end state or cyclical return: For linear processes, state the final output. For cycles, note that the final stage loops back to the start.

STAGE Applied: Linear Process

Hypothetical image: "How Paper is Made" — 6 stages: 1. Trees felled → 2. Logs transported → 3. Wood chipped → 4. Pulp produced → 5. Paper sheets formed → 6. Paper dried and cut.

STAGE response:

"The diagram illustrates the industrial process by which wood is converted into paper, comprising six sequential stages. The process begins when trees are felled and the logs are transported to a processing facility. The logs are then chipped into small pieces, which are converted into wood pulp through a chemical or mechanical process. The pulp is subsequently spread onto large sheets and pressed to form paper, which in the final stage is dried and cut into standard dimensions. The overall process transforms raw timber into a standardised commercial product through a continuous, linear sequence."

Word count: ~105 words | Time: ~40 seconds | Content score: 4.5–5.0

STAGE Applied: Cyclical Process

Hypothetical image: "The Water Cycle" — 5 stages: Evaporation → Condensation → Precipitation → Surface Runoff → Collection → (back to Evaporation).

Key adaptation for cycles: The A step states that this is a cyclical process. The E step explicitly names the return.

"The diagram illustrates the water cycle, a continuous natural process comprising five interconnected stages. Beginning with evaporation, water from oceans and lakes is converted to vapour by solar energy and rises into the atmosphere. Cooling causes the vapour to condense, forming clouds, which subsequently release water as precipitation — rain, snow, or sleet. Surface runoff channels this water into rivers and water bodies, where it is collected before evaporating again. The process is cyclical, with no fixed endpoint, as collection continuously feeds back into the evaporation stage."

What's different for cycles: "continuous," "no fixed endpoint," and the explicit return statement in the E step. These three cues signal to the AI that you understand the cyclical nature — which is the core Content expectation for this image type.

Common Process Diagram Mistakes

Mistake 1: Passive enumeration. "First there is evaporation. Then there is condensation. Then there is precipitation." This scores Content 2.5–3.0. The AI rewards causal connectors: "evaporation produces vapour, which rises and condenses." The relationship between stages is what the rubric measures.

Mistake 2: Describing the arrows. "There is an arrow pointing from evaporation to condensation." Arrows are structural notation — describing them wastes time and earns no Content credit. Describe what the arrow means (the causal relationship), not the arrow itself.

Mistake 3: Treating a cycle as a line. Ending a cyclical process description with "and this is the final step" misses the return-to-start logic. Always close cycles with the return statement.

Mistake 4: Inventing detail. For process diagrams, the AI does not reward elaboration beyond what the image shows. If the diagram labels a stage as "pulp production" without further detail, describe it as "conversion of wood chips into pulp" — do not add information about the chemical reactions involved. Fabrication does not improve Content score and risks coherence penalties.

Framework 3: LOCATE — PTE Describe Image Maps

Maps are the least practiced DI image type among PTE students — and often the most rewarding, because the spatial-language framework is learnable in under a week. Map tasks in PTE DI come in three forms: single reference maps (a map with labeled features), before/after comparison maps (same area, two time points), and route maps (a path from A to B with landmarks).

L — Label the geographic subject: State what region, city, or area the map depicts. Include the stated purpose if one is given (e.g., "a proposed urban development," "infrastructure changes between 1990 and 2020").

O — Orient the dominant feature: Identify and locate the most important feature in spatial terms. Use directional language: north, south, east, west, central, coastal, inland, adjacent to.

C — Compare or contrast: For before/after maps — what has changed? For single maps — what is the relationship between the major features (proximity, spatial pattern, distribution)? For route maps — what is the path's orientation and what landmarks does it pass?

A — Additional features: Name 1–2 secondary features that add context or contrast.

T — Takeaway pattern: Close with the spatial logic — is development concentrated (in one area), dispersed (spread across the map), linear (along a road or river), or radiating (expanding outward from a center)?

E — Express the overall conclusion: One sentence summarising what the map communicates as a whole.

LOCATE Applied: Before/After Map

Hypothetical image: Two maps of "Greenfield Town, 1990 and 2020." 1990: farmland to the north, small town center, single road. 2020: residential areas replaced farmland, two new roads, shopping center added east of town center.

LOCATE response:

"The maps compare the urban development of Greenfield Town between 1990 and 2020. In 1990, the town centre was the dominant feature, situated centrally and accessed by a single road, with agricultural land occupying the northern zone. By 2020, the northern farmland had been replaced by residential development extending outward from the town centre, and a new shopping centre had been added to the east. A second road now connects the expanded residential area to the town centre. The overall pattern shows significant outward growth — particularly northward and eastward — reflecting a clear trend of urban expansion over the three-decade period."

Word count: ~110 words | Time: ~42 seconds | Content score: 4.5–5.0

(Note: aim for ~105 words to stay safely within 40 seconds at a moderate speaking pace.)

LOCATE Applied: Single Reference Map

Hypothetical image: Map of a university campus showing: library (central), lecture halls (northern block), cafeteria (south), sports complex (east), student accommodation (west and northwest).

LOCATE response:

"The map shows the layout of a university campus, with the library positioned centrally as the dominant facility. Lecture halls occupy the northern section of the campus, adjacent to the main academic zone, while student accommodation is distributed across the western and northwestern areas — providing proximity to both academic and social facilities. A sports complex is located to the east, and the cafeteria occupies the southern section. The campus layout follows a hub-and-spoke pattern, with the central library accessible from all four residential and academic zones."

Word count: ~95 words | Time: ~37–38 seconds | Content score: 4.0–5.0

Spatial Language Reference: Essential Vocabulary

Directional: north of, south of, east of, west of, central, peripheral, coastal, inland, adjacent to, opposite, at the intersection of

Relationship: bordered by, separated from, connected to, accessible from, overlooking, situated between

Change (for before/after): replaced by, converted to, expanded northward, extended along, demolished and rebuilt, newly constructed, remained unchanged

Distribution: clustered in, dispersed across, concentrated along, linear arrangement, radiating outward from

Cross-Image-Type Timing Strategy

All three frameworks share the same 40-second architecture:

| Segment | Target Duration | Word Count | |---------|----------------|------------| | Opening statement (subject + scope) | 6–7 seconds | ~16–20 words | | Primary feature / dominant element | 9–10 seconds | ~24–28 words | | Comparison / secondary element | 11–12 seconds | ~28–32 words | | Closing synthesis | 8–9 seconds | ~20–24 words | | Total | ~38–40 seconds | ~88–104 words |

The 38-second rule: Aim for 38 seconds, not 40. A 40-second response that clips the final word sounds rushed and damages Oral Fluency. A 38-second response with a clean closing sentence sounds controlled and deliberate — and scores equally on Content while protecting Fluency.

4-Week Practice Plan: All Three Image Types

Week 1: Single-Type Identification & Framework Drilling

Daily (35 minutes):

• Days 1–2: Pie charts only. Use SLICE on 5 practice images per day. Record yourself and check: did you name exactly 3 segments, use a fraction or comparative expression, and close with a synthesis?
• Days 3–4: Process diagrams only. Use STAGE on 5 practice images. Check: did you state the step count, use causal connectors, and identify whether it's linear or cyclical?
• Days 5–7: Maps only. Use LOCATE on 5 practice images. Check: did you use directional language in every sentence about feature location?

Weekly target: Consistent framework use — opening sentence names image type correctly for 90% of images.

Week 2: Comparison Practice

Daily (40 minutes):

• Specifically target double pie charts and before/after maps — the comparison variants that require "change language"
• For each double pie, identify the single biggest shift and build the I and C steps around it
• For each before/after map, list the 3 changes before speaking — choose the 2 most significant to include
• Record 6 responses per day; review for: change-language accuracy (did you say "increased from X to Y" or "replaced"), and closing synthesis quality

Weekly target: Change-language present in 100% of comparison-type responses.

Week 3: Mixed Type Randomisation

Daily (45 minutes):

• Mix all six subtypes (single pie, double pie, linear process, cyclical process, single map, before/after map)
• Do NOT look at what type is coming — identify the type in the first 5 seconds and select framework automatically
• Time yourself: flag any response that exceeds 40 seconds or falls below 30 seconds
• Focus fix: the 30-second responses are almost always missing the closing synthesis — isolate and add it

Weekly target: Correct framework selection for 95% of images; all responses between 35–40 seconds.

Week 4: Full Mock Integration

Daily (50 minutes):

• Complete 2 full PTE Speaking section mocks — all DI tasks included (typically 6–7 per test)
• After each mock, categorise DI errors by type: wrong framework, missing synthesis, timing error, content fabrication
• Spend 20 minutes on targeted repair: if your pie chart responses are weakest, drill 5 more SLICE responses that day
• Submit 2 recorded responses to a qualified PTE coach for Content score feedback

Weekly target: Consistent 4.0+ Content on all three image types in mock conditions.

Related Posts

This post sits within KS Institute's PTE Speaking topic cluster. For complete preparation, combine this guide with:

• PTE Describe Image: Advanced Data Narration for Multi-Trend Line Graphs & Comparative Bar Charts (2026) — the paired guide covering the data chart types this post does not address
• PTE Read Aloud: Prosodic Chunking & Pause Placement for 79+ Oral Fluency (2026) — Oral Fluency techniques that complement DI delivery
• PTE Re-tell Lecture Advanced: Complex Lecture Structures & 79+ Content Score (2026) — the advanced guide for the other image-inclusive Speaking task

Frequently Asked Questions

Q: My DI tasks only seem to have bar charts and line graphs. Do pie charts and maps actually appear in live PTE tests?

Yes — pie charts, process diagrams, and maps are confirmed item types in PTE Academic. The mix you encounter depends on the item bank rotation for your test session. Students who report "never seeing" these types typically encounter them on retakes. Leaving them unprepared is a significant risk for students targeting 79+.

Q: For process diagrams, do I need to name every stage?

No. The Content rubric rewards logical completeness, not exhaustive enumeration. For a 6-stage process, cover 4–5 stages using causal connectors. The stage you skip should be a secondary transition step, not the starting point, major transformation, or final output.

Q: Is there a penalty for using the wrong framework on an image type?

Not explicitly — the PTE AI does not know which framework you used. What it scores is whether your response correctly communicates the relationships in the image. Using a data-trend template on a process diagram will fail because you'll produce comparison language that doesn't match the sequential content — and the AI will score that as low Content accuracy.

Q: For maps, how specific do I need to be about compass directions?

Approximate directional language is fine and preferred: "the northern section," "to the east of the town centre," "along the western boundary." You do not need to say "45 degrees northeast" or give precise geographic coordinates. The goal is spatial clarity, not cartographic precision.

Q: How does Describe Image scoring work — is it all one score or broken into components?

PTE Describe Image is scored across three components: Content (0–5), Oral Fluency (0–5), and Pronunciation (0–5). The frameworks in this guide target Content. Oral Fluency is primarily improved through connected speech and pause placement — see our Read Aloud guide for techniques. Pronunciation is addressed in our PTE Speaking Pronunciation guide.

Q: I consistently score well on Content but my overall Speaking is still 68–72. What else is affecting my score?

At KS Institute, we find that students with strong Content scores but Speaking below 75 typically have Oral Fluency issues — specifically mid-response hesitations, uneven pacing between tasks, and over-long silences when transitioning from notes to speech. Book a free Speaking diagnostic with our coaches to identify your specific bottleneck: we'll listen to 3 of your DI recordings and score each component individually.

Q: Can I use these frameworks in IELTS as well?

The SLICE framework is directly applicable to IELTS Academic Writing Task 1 pie chart descriptions (as a written rather than spoken framework). The LOCATE framework applies to IELTS Task 1 map questions. STAGE is directly applicable to IELTS process diagram tasks. PTE students preparing for IELTS retakes can transfer these frameworks with minimal modification.

The Bottom Line

PTE Describe Image — across all image types — is a relationship communication task, not a feature-listing task. The PTE AI doesn't reward you for naming more elements; it rewards you for showing that you understand how the elements of the image relate to each other.

The SLICE, STAGE, and LOCATE frameworks each encode the specific relationships that the AI measures for their respective image types:

• SLICE encodes proportional comparison and distributional logic (pie charts)
• STAGE encodes causal sequence and process closure (process diagrams)
• LOCATE encodes spatial relationship and change pattern (maps)

These are not templates you fill in mechanically. They are cognitive structures that guide you to say the right kind of thing for each image type — which is what separates Content 4.5–5.0 from Content 2.5–3.0.

KS Institute has prepared 5,000+ students for PTE Academic over 19 years. Gagan Daga, who leads our Speaking programme with 15+ years of PTE coaching experience, developed the SLICE, STAGE, and LOCATE frameworks based on analysis of high-scoring and low-scoring DI responses across our student cohort. 82% of KS Institute students targeting 79+ in Speaking achieve it within 8 weeks — with Describe Image improvement being one of the highest-leverage interventions.

Ready to get your DI responses scored by a PTE expert? Book a free 20-minute Speaking diagnostic and we'll score your SLICE, STAGE, and LOCATE responses component by component. Book your free assessment →