Part 1 — Introduction and interview (4–5 minutes)
Answer each question aloud in 2–4 natural sentences. Press 🔊 to hear the examiner ask it. Don't script — react.
Part 2 — Individual long turn (3–4 minutes)
You have one minute to prepare (make notes!) and should then speak for one to two minutes. Use the timers — the discipline is the practice.
Describe a memorable conversation you had with a stranger.
You should say:
- who this person was and where you met them
- how the conversation started
- what you talked about
- how long you spent talking
and explain why this conversation has stayed with you, and whether it changed how you feel about talking to people you don't know.
You will have one minute to prepare and should speak for one to two minutes.
Part 3 — Two-way discussion (4–5 minutes)
These get harder as they go. Build answers with: Answer → Explain → Example → Alternative view → Evaluation. Open the follow-ups only after you've answered — that's what the examiner asks when an answer is thin.
Examiner follow-ups
- What is it about those situations that gives people permission?
- Is it the place, or the fact that you'll probably never meet again?
Examiner follow-ups
- Does that depend on who is doing the talking — their age, their gender?
- Has that attitude changed within your lifetime?
Examiner follow-ups
- Could a society function without it?
- Why do you think some people are so bad at it?
Examiner follow-ups
- Would city life work if everyone greeted everyone?
- Does that anonymity affect how people treat each other?
Examiner follow-ups
- Do you think that childhood rule makes adults more hesitant?
- How else could we teach children to be careful?
Examiner follow-ups
- Are the connections apps create a fair replacement?
- Is avoiding strangers a choice people are entitled to make?
Examiner follow-ups
- What's the difference between trusting a person and trusting a platform?
- Could that kind of system-based trust ever go badly wrong?
Examiner follow-ups
- What are people actually afraid will happen?
- Is that fear realistic, in your experience?
Examiner follow-ups
- Can a government do anything about it, or does it only grow from the bottom up?
- How quickly can that trust be lost?
Examiner follow-ups
- Is there a skill being lost, or just a habit?
- Would anything realistically reverse that trend?
What a Band 8–9 performance should demonstrate
Judge your recording against the four official criteria. An accent is not required — clarity is.
Fluency & Coherence
- Tells the Part 2 story with momentum — how it started, what was said, why it stuck — without circling back or restarting sentences.
- Reconstructs dialogue smoothly ("so she asked me why I was carrying a telescope, and I ended up explaining…") — reported speech without losing the thread.
- In Part 3, takes on genuinely hard questions (trust, anonymity) at a steady pace instead of speeding up nervously or grinding to a halt.
- Uses natural hesitation devices when thinking ("Honestly, I've never been asked that — my instinct is…") rather than dead air.
- Connects answers to each other: the point about headphones in one answer gets picked up when the rating-system question arrives.
Lexical Resource
- Precise topic language: strike up a conversation, break the ice, keep to yourself, a passing acquaintance, common ground, guard down.
- Distinguishes near-synonyms cleanly: stranger vs outsider vs foreigner — mixing these up matters on exactly this topic.
- Abstract Part 3 vocabulary handled naturally: anonymity, social fabric, vulnerability, etiquette, accountability, at face value.
- Idioms placed sparingly and correctly: "we just clicked", "it broke the ice" — one apt idiom per section, not a shower of them.
- Paraphrases 'talk to strangers' many ways across the test: approach someone, strike up a chat, engage with people you don't know, make small talk.
Grammatical Range & Accuracy
- Past narrative machinery in Part 2: "I'd been waiting for about an hour when the man next to me asked…" — past perfect and continuous doing real work.
- Speculation about people in general: "people tend to…", "most of us would rather…", "there must be something about trains that…".
- Conditionals for the sociological questions: "If everyone greeted everyone in Mumbai, nobody would get anywhere."
- Contrast structures: "Whereas in a village everyone knows your business, in a city nobody knows your name — and both cut both ways."
- Complex noun phrases without collapse: "the kind of trust that lets you fall asleep on a train full of people you've never met".
Pronunciation
- Two voices in Part 2 — your narration and the stranger's reported words — separated by subtle shifts in tone, which is high-level delivery.
- Stress lands on the operative contrasts: "we trust the platform, not the person".
- Question-like intonation used deliberately when wondering aloud ("and maybe that's the point?") — variety that sounds thought, not read.
- Connected speech in high-frequency chunks: "sort of", "you know what I mean", "never met them before" flow as single units.
- Polysyllabic abstractions — anonymity, spontaneous, vulnerability, acquaintance — stressed correctly even mid-argument.
How to develop a Band 9 answer for this test
- Set the scene fast — where you were, why you were there, and one visual detail (the delayed train, the rain, the empty waiting room) so the stranger enters a real place.
- Show the opening move — who spoke first and the actual first line if you remember it; the mechanics of how strangers start talking is exactly what this card wants.
- Compress the middle — pick the two or three turns the conversation took rather than replaying all of it; "we went from cricket to his years in the merchant navy" covers ground quickly.
- Mark the shift — the moment it stopped being small talk: the surprising confession, the shared coincidence, the advice you didn't ask for but needed.
- Answer the time bullet honestly — twenty minutes or four hours, and one clause on how it felt ("it went in a blink") to keep it a story, not a log.
- Close on the residue — why it stayed with you and whether it changed your behaviour toward strangers; a small admission ("I still look for a seat next to someone interesting") lands the explain-clause.
“We routinely get into strangers' cars and sleep in strangers' homes through apps. Has technology actually increased trust between strangers, or just outsourced it to a rating system?”
- Answer: commit to the sharper reading — e.g. it's mostly outsourcing: we haven't learned to trust strangers, we've learned to trust the score attached to them.
- Explain: unpack the mechanism — the platform absorbs the risk assessment we used to do ourselves; the driver is a 4.9, so no judgement of the human is required, and the skill of reading people atrophies.
- Example: make it concrete — you'll get into an unknown driver's car at midnight because an app says so, but might hesitate to accept a lift from a neighbour of ten years; the trust follows the interface, not the person.
- Alternative view: concede what's real — repeated safe experiences do generalise; a generation that has couch-surfed across Europe is demonstrably more comfortable with strangers than their parents, whatever the mechanism.
- Evaluate: land it — call it borrowed trust: it works brilliantly until the system fails, and a society should notice when its confidence in strangers can be switched off by a server outage.
Common mistakes for this specific test
Candidates reach for their best conversation story, which is usually with someone they knew. The card's engine is unfamiliarity — the not-knowing is what makes it memorable. If the person was already a friend, the story misses the point of every bullet.
"We talked about many things and it was very nice" gives the examiner nothing. One line of reported speech — what they asked, what you answered — is the difference between a story and a summary.
A suspiciously perfect stranger who delivers life-changing wisdom on a mountaintop sounds scripted, and scripted delivery follows. A real, modest encounter — the aunty on the bus who talked for an hour — is easier to speak about naturally and scores better.
The questions probe trust, culture and cities. Answering each one with warnings about crime and being careful flattens ten different questions into one repeated safety announcement.
Saying people in your country never talk to strangers, then describing constant train-compartment friendships two answers later, signals answers assembled from stock phrases rather than thought. Keep your own picture consistent.
The topic traps you into repetition. Rotate: people you don't know, a passer-by, a fellow passenger, someone new, an unfamiliar face. Range here is cheap and visible.
That question hands you a paradox — we avoid what's good for us — and asks you to explain it. Agreeing with the research and stopping is a Band 6 move; the marks are in the why: fear of rejection, scripts we lack, energy costs.
"Cities are cold and villages are warm" is a cliché the question explicitly challenges — it offers anonymity as possibly what makes city life bearable. Engaging with that provocation, even to reject it, shows the flexibility Band 8–9 needs.
"I use social media and messaging apps" describes every phone on Earth. Name the app, name the habit — the examiner is scoring the specificity of your language, not the sophistication of your phone.
The later Part 3 questions are the hardest in the test, and nervous candidates sprint through them in half-sentences. A slower, structured answer to one hard question outscores three breathless generalisations.