From one sentence to one minute: building longer speaking answers
A sentence you can say cleanly often falls apart across a full minute of unscripted speech. A staged way to build from single sentences to sustained answers.
30-second version: Speaking one clean sentence and sustaining a full, coherent minute of unscripted speech are different skills — vocabulary and pronunciation that hold up in a single rehearsed line often degrade by the third sentence of a real answer, under the combined load of content-planning and delivery. The fix is staged: build from a single sentence, to a few linked sentences, to a full minute, adding real-time pressure gradually rather than all at once.
A pattern worth naming honestly: a learner can read a single sentence aloud cleanly, with good pronunciation and pace, and then completely lose that same quality three sentences into an unscripted one-minute answer — more filler words, worse pacing, words half-remembered instead of said with confidence. Nothing about their vocabulary or pronunciation actually got worse between the sentence and the paragraph. What changed is the load: constructing content on the fly while also delivering it cleanly is a genuinely harder combined task than delivering pre-existing text, and most practice never builds up to that combined load deliberately.
Why length specifically breaks things
A single rehearsed or read sentence asks for delivery only — you already know what you're going to say. A full unscripted answer asks for delivery and content-planning simultaneously, in real time, and the two compete for the same attention. Vocabulary retrieval that's instant when you're reading a script becomes noticeably slower when you're also deciding what to say next, which is exactly where fillers, hesitation, and reaching for a simpler word than you'd planned all creep back in.
A staged routine
Stage 1 — One clean sentence. Read or say a single sentence, focusing purely on delivery: accuracy, pace, no fillers. This isolates the delivery skill from the content-planning load entirely.
Stage 2 — Two to three linked sentences, lightly planned. Give yourself ten seconds to think of two or three sentences on a simple prompt ("describe your morning"), then say them together as one continuous answer. This adds a small amount of real-time content decisions while keeping the total load manageable.
Stage 3 — Thirty seconds, unscripted. Pick a prompt and speak for thirty seconds with no preparation time at all. Some disfluency here is expected and fine — the goal isn't a clean take, it's building tolerance for speaking while still deciding what to say.
Stage 4 — A full minute, unscripted. The target length for most real speaking situations — an interview answer, an IELTS Part 2 response, an explanation to a colleague. A full minute genuinely tests sustained fluency in a way thirty seconds doesn't, because the first thirty seconds of adrenaline-assisted focus has worn off and you're relying on the underlying habit, not a burst of initial effort.
What to actually check at each stage
Don't just listen back and feel good or bad about it — check the concrete numbers. Your fluency score, filler count, and pace should be roughly consistent between a rehearsed sentence and stage 4's unscripted minute — if they degrade sharply as length and spontaneity increase, that's the specific gap this staged practice is meant to close, and it's diagnostic: it tells you the underlying skills are there, they just haven't been tested under full real-time load yet.
Where new vocabulary fits into this
Deliberately using recently learned vocabulary inside a longer unscripted answer is the real test of whether it's usable, not just known — see why definitions alone don't make a word usable for the case that retrieval-under-pressure is a separate skill from definition or even isolated pronunciation. A minute of unscripted speech is exactly the pressure that skill needs to be tested against.
A weekly target
Once a week, run the full staged sequence on a new prompt, and once a week, skip straight to stage 4 on a prompt you haven't prepared at all — the second version is the one that actually tells you where you stand.
Try it: pick a simple prompt right now, give yourself no preparation time, and speak for a full minute — then check your filler count and pace against a rehearsed sentence from earlier today.
Hear exactly which sounds to fix.
Say one sentence and get sound-by-sound feedback in seconds. No install, no card.