Almost every learner has the same experience: they understand the teacher, the course audio and the news presenter, then a real person speaks and they catch one word in five.
This is not a vocabulary problem. It is that real speech is a different signal from classroom speech, and most listening practice never trains for it.
Why real speech is harder
Classroom English gives you every word, fully pronounced, at a steady speed. Real English does three things to that signal:
- It compresses. Going to becomes gonna; what do you becomes whaddaya; him loses its h. Small words are squeezed until they are barely there.
- It links. An apple sounds like a napple; pick it up runs into one word.
- It varies. Speakers hesitate, restart, change direction mid-sentence and use rhythm, not grammar, to show what matters.
You can know every word in I should've told her about it and not recognise it at speed, because it does not sound like six words. It sounds like two.
The fix is exposure to exactly this kind of speech, in doses small enough to work on.
What to listen to at each stage
| Stage | Material | Length | Why |
|---|---|---|---|
| Starting out | Slow, clear speech: graded podcasts, learner news | 1–2 min | Build the base without drowning |
| Comfortable with clear speech | Interviews, everyday conversations, vlogs | 2–3 min | Natural speed, real hesitation |
| Losing real conversations | Unscripted chat: casual podcasts, group discussions | 1–2 min clips | Compression and overlap |
| Preparing for a specific situation | Meetings, calls, check-ins for your goal | Short excerpts | Train the exact patterns you will meet |
Short is the important word. A ninety-minute podcast is entertainment. A two-minute clip you listen to four times is practice.
The routine: one clip, three passes
Take a clip of one to two minutes. Do not use a transcript yet.
Pass 1 — gist. Listen once. Say out loud, in one sentence, what it was about. Do not worry about details.
Pass 2 — detail. Listen again with one question in mind: what exactly did they say about X? Write down the phrase as you heard it, even if it is wrong.
Pass 3 — shadow. Pick three sentences and repeat each one straight after the speaker, copying the rhythm and the compressions. This is where listening becomes pronunciation, and where compressed forms start to sound normal.
Now check with a transcript or subtitles. Look only at the places you got wrong. Those are usually compressions and links, not unknown words — which is the whole point.
Ten to fifteen minutes. Every day. Same clip two days in a row is fine; a clip you finally understand is more valuable than a new one you do not.
Predict before you listen
Before pressing play, spend thirty seconds guessing: who is speaking, about what, which words are likely to appear. Prediction is what native listeners do constantly, and it is why they can follow speech with half the words missing. Learners who listen with an empty mind have to catch every word; learners who predict only have to confirm.
Listening inside a conversation
The hardest listening is the kind where you also have to answer. Two habits help:
Have clarification phrases ready, so a missed word does not become a stopped conversation:
Sorry, could you say that again? / Do you mean…? / Just to check — you said Tuesday?
Confirm before you answer. Repeating the key point back — so you want it by Friday? — buys time, checks understanding and sounds professional. Nobody minds.
For how to set up conversations where you get to practise this, see English conversation practice online.
Listening and pronunciation are one skill
You cannot reliably hear a contrast you cannot produce, and once you can produce it you start hearing it everywhere. That is why the shadowing pass matters: every compressed form you can say becomes one you can catch. The pronunciation practice guide covers the production side with the same short-clip routine.
Where iTellU fits
Listening improves fastest when it is followed immediately by speaking. In the iTellU app, short listening and speaking practice sit next to each other, and live one-to-one lessons with real teachers give you unscripted speech — hesitations, follow-ups, compressions — that no recording reproduces. Ask your teacher to speak at natural speed and to let you repeat back what you heard before you answer.
Downloading the app is free; lesson plans are on the pricing page.
FAQ
- How much listening practice should I do each day?
- Ten to fifteen minutes of active listening — short clips, repeated, with a task — does more than an hour of background audio. Add passive listening on top if you enjoy it, but do not count it as practice.
- Should I use subtitles?
- Use them in the right order: first listen without, then with subtitles in English to check, then without again. Watching with subtitles on from the start trains reading, not listening. Subtitles in your own language train translation.
- Why do I understand my teacher but not people in films?
- Teachers speak clearly, slowly and with full forms. Real speech is faster, compresses small words and runs them together — 'what do you want to' becomes 'whaddaya wanna'. You need practice with that kind of speech specifically, in short doses.
- Should I listen to things I don't understand?
- Mostly no. Listening you understand almost none of teaches little; aim for material where you catch most of it and miss some. The exception is short shadowing clips, where the goal is copying rhythm rather than meaning.