The Skill Profile is the most specific, most actionable thing Pearson gives you — and most candidates never read it. This is the complete breakdown: all 8 categories, every task weighting, the August 2025 changes, a real case study, and a 4-week plan to fix exactly what's holding your score back.
Sample Skill Profile
Shortest bars = highest priority. They point to the exact tasks to drill before your next exam.
A Speaking score of 36 doesn't tell you whether the problem is fluency on long responses, accuracy on short responses, connected-speech reproduction, or content under time pressure. Your Skill Profile does. Once you know which specific bars are low, you know — with precision — which of the 22 task types to drill, and in which order.
This page gives you the complete picture: every Skill Profile category, every question type, every common mistake, and every fix.
Summarize Group Discussion (20% Listening, 19% Speaking) and Respond to a Situation (13% Speaking) became fully scored on 7 August 2025 after 15 months as trial items.
Since 2023, Pearson's AI weights content recognition far more heavily. Memorised templates and generic phrases that used to pass are now detected and scored lower.
For Describe Image, Retell Lecture, SGD, RTS, SWT, Essay and SST, ambiguous or off-topic responses are now routed to a human. You cannot game these with templates.
Third-party practice portals are not aligned to Pearson — often not even 10%. Get a manual qualitative assessment of speaking and writing before your next exam.
Why scores drop
Describe Image with no formula runs out of content before 40s (a fluency failure). Essays that describe instead of committing to a thesis. SGD treated like Retell Lecture. SST compressed from two directions by weak listening and weak writing.
How to fix it
Describe Image: four-part formula every time (Open → Main Point → Comparison → Conclude), 4 anchor words in the 25s prep, 70–80 words. Essay: thesis in line one, PEEL each paragraph, 220–260 words. SGD: track themes not speakers. SST: use the full 10 minutes, content first then grammar, 50–70 words.
Three tasks — Describe Image, SGD, Repeat Sentence — are 66% of your Speaking score.
Summarize Group Discussion is the single highest-weight Listening task — and the most under-practised.
Summarize Written Text sits in the Speaking & Writing section yet contributes 23% of your Reading score.
Write from Dictation contributes 23% of Writing — yet it is a Listening-section task.
One hour on Repeat Sentence moves two score columns and three Skill Profile bars. One hour on Multiple Choice Single Answer moves one column and one bar. Always address cross-skill tasks first.
Portal scoring is not Pearson scoring. Those encouraging numbers gave a false sense of readiness — a 40-point gap from her best portal score to the real exam.
Her Skill Profile bands — shortest = highest priority
Short Speaking (Read Aloud + Repeat Sentence), Reproducing (adds Write from Dictation) and Short Writing (Write from Dictation only) are all below average. From the Skill Profile alone — before seeing any individual question score — we can identify with high confidence that Repeat Sentence and Write from Dictation are her two most underperforming tasks.
Two categories is the practical working limit for four weeks of focused preparation. Don't try to fix all eight at once.
Short Writing low → Write from Dictation only. Extended Speaking low → Describe Image, Retell Lecture, SGD, Respond to a Situation.
Any task on both lists is your absolute first priority — one improvement moves both bars. If no overlap, fix the higher-percentage task first.
Moved up → the drill works. Didn't move → adjust or give it more time. Dropped → a new problem; investigate before the next cycle.
Tier 1 — Cross-skill multipliers (first)
Repeat Sentence · Write from Dictation · Summarize Group Discussion · Retell Lecture · Summarize Spoken Text
Tier 2 — High-weight single-skill tasks
Describe Image (S 31%) · Fill in the Blanks Dropdown (R 25%) · Write Essay (W 31%)
Tier 3 — Medium-weight tasks
Summarize Written Text · FIB Drag & Drop · Respond to a Situation · Read Aloud · Reorder Paragraph · Highlight Incorrect Words
Tier 4 — Low-weight tasks (maintain)
All MCQ variants · Highlight Correct Summary · Select Missing Word · Answer Short Question · FIB Type In
Identify your two lowest bars, map them to question types, and build a targeted drill schedule.
Have someone who understands what the AI evaluates review your spoken and written responses.
Not general practice. Task-specific practice produces task-specific improvement.
If a bar didn't move, drill one more week. If a new bar dropped, fix it before sitting.
In a free 30-minute one-on-one session we identify exactly what went wrong, pinpoint your two lowest bars, and map a targeted drill schedule to your exam date. No charge, no obligation.
Prefer to read the full long-form version? Read the complete Skill Profile guide on our blog.