The card rewards process. What you tried and abandoned is worth more than what finally worked.
The cue card
Describe a problem you solved.
You should say
1 minute to prepare · up to 2 minutes to speak
Four beats, not four answers. The prompts on the card are a scaffold — a turn that flows past them scores better than four disconnected replies.
Lexical resource is a quarter of your band, and band 7 explicitly allows the occasional wrong word — playing it safe with plain vocabulary is what keeps people at 6.
get to the bottom of
for diagnosing a cause
a stumbling block
for the specific thing blocking progress
work around
for solving without fixing
it dawned on me
for a realisation arriving gradually
trial and error
for iterative problem-solving
nip it in the bud
for catching a problem early
The failed attempt is the best material on the card. Include it — narrative with a setback in it is easier to speak at length about.
Read this to see the moves, not to learn the words. An examiner hears memorised answers every working day and it shows up in your fluency score — the hesitation pattern of someone reciting is nothing like the pattern of someone thinking. The details below are invented; yours should be your own.
The problem I'd describe is a fairly mundane one at work: our team's weekly report took someone an entire day to produce and nobody could remember why. I'd been there about four months, which is the useful window where you can still ask stupid questions. So I asked who read it. It turned out the answer was essentially nobody — it had been created for a manager who'd left eighteen months earlier, and it was still being produced every week because it always had been. Solving it wasn't technical at all. I spent two afternoons tracing who actually used which section, found that three people wanted three specific numbers, and replaced the whole thing with an email that takes about ten minutes. What I took from it is that a lot of problems aren't hard, they're just unexamined, and the thing that stops people examining them is the assumption that somebody must have already had a reason. Often nobody has.
Part 3 is loosely tied to your long turn, but the questions turn abstract — about people in general rather than about you. This card usually leads into:
The word is journey, not destination — and the difference is worth a band.
A card built on a fork in the road. If there was no real alternative, pick a different decision.
Cultural detail is free marks here — the examiner may genuinely not know how the occasion works.
A narrative with built-in tension — one of the easiest cards to speak fluently on if you tell it in order.
Mirror image of the person-who-helped-you card, and the same rule applies: build the problem first.
Needs a turn in it. A mistake with no consequence and no lesson is just an anecdote.
Oliver will read you a card, wait out your preparation minute, and then say nothing at all for two minutes — which is the part you cannot rehearse alone. The report afterwards scores fluency, vocabulary, grammar and pronunciation separately. The first test is free.
This guide is written from the publicly published IELTS Speaking band descriptors, in our own words. For the official wording, see the British Council's band descriptors. MockLive AI is an independent practice tool and is not affiliated with IELTS, the British Council, IDP or Cambridge Assessment English.