Voice: Volume, Pace and Being Understood

Here is a thing almost nobody prepares for. You can build excellent answers, structure them well, know the ethical frameworks cold, and still lose marks because the person marking you had to work to hear you.

Voice is not a soft extra. It is the delivery mechanism for everything else, and it degrades in exactly the conditions an interview creates: adrenaline, an unfamiliar room, a stranger two metres away, sometimes a mask or a screen in between. The fixes are mechanical and they work quickly.

Volume: almost everyone is too quiet

Nerves shrink the voice. Your body tightens, your breathing moves up into your chest, and your speaking volume drops without you noticing, because your own voice always sounds loud from inside your head. Meanwhile the room is working against you: hard surfaces, air conditioning, a corridor full of candidates, an assessor sitting further away than a friend would.

If masks are in use at your interview, add another layer of loss on top, and remember that the assessor also loses your mouth as a visual aid. Anything that muffles you makes volume and articulation matter more, not less.

The correction is uncomfortable and simple: speak at a level that feels slightly too loud to you. That level is almost always normal to the listener. If you would be comfortable being overheard by someone at the next table in a cafe, you are roughly right.

Volume comes from breath, not from strain

Trying to be louder by pushing from the throat gives you a tight, slightly aggressive sound and a sore voice by station six. Volume properly comes from air.

Two adjustments. First, breathe low: let the air go into your belly rather than lifting your shoulders. You can check by putting a hand just below your ribs and feeling it move outward as you inhale. Second, aim at the far wall rather than at the assessor's face. Directing your voice past them, mentally, produces the right projection without any sense of shouting.

Posture does the rest. Sitting back with a collapsed chest compresses your breathing. Sit forward slightly, feet on the floor, spine long. It is not about looking keen. It is about having air available.

Finish your sentences

The single most common vocal fault in interviews is not being quiet throughout. It is starting each sentence at a reasonable volume and losing the last four words as your breath runs out. Those last four words are usually where your actual conclusion lives.

Two habits fix it. Keep sentences shorter, so you never run out of air before the full stop. And consciously give the end of the sentence the same energy as the start, particularly the final word, which many people drop to almost nothing.

Related: let statements sound like statements. When every sentence rises at the end, a considered position starts to sound like a series of questions you are hoping the assessor will approve of. You are allowed to sound certain about things you are certain about. Land the last word and stop.

Pace, and the pauses that make you sound thoughtful

Speed is the enemy of being scored well, because an assessor writing notes while you talk needs processing room. When you understand how MMI scoring works in Australia, the reason becomes obvious: they are mapping what you say onto specific domains, in real time, and a wall of fast speech is much harder to map than a paced answer with visible structure.

Deliberate pauses do three jobs at once. They give the assessor time to catch up and write. They give you time to plan the next sentence rather than filling the gap with noise. And they read as considered rather than blank, provided your face is doing something other than panic.

Put your pauses at the joints of the answer: after you state your structure, between your main points, and before your conclusion. A pause in a sensible place sounds like punctuation. A pause in the middle of a clause sounds like you lost your thread.

Warm up before you go in

Singers and actors would never perform cold, and a circuit is a performance in the physical sense at least. A cold voice at station one is a real disadvantage, and station one still counts. Twenty minutes before, somewhere private:

• Talk out loud for a few minutes, to anyone or to yourself in the car. Silence for an hour before an interview leaves you cold and quiet.
• Hum gently up and down through your range to get the resonance working without straining anything.
• Over articulate a couple of sentences, exaggerating the consonants, then return to normal. Your normal will be crisper.
• Roll your shoulders and unclench your jaw. Tension in the neck and jaw directly narrows your sound.
• Drink water, at room temperature, and keep drinking through the day. A dry throat at station five is avoidable.
• Skip the heavy coffee if it makes your hands shake. Caffeine tightens some people's voices considerably.

Online changes the physics

A microphone compresses your dynamic range, so the difference between your loud and quiet sounds flattens, and small pauses feel much longer to both sides because of latency. Our comparison of virtual and in person MMIs goes through the wider differences, but vocally the rules shift in three ways.

Speak slightly more slowly than you would in person, since any lag eats your first syllables. Do not lean in toward the microphone when you get intense, because that spikes the level unpleasantly. And add a beat of silence after the assessor finishes before you begin, so you are not clipping each other. Test your audio on the actual platform, with the actual headset, before the day.

Roleplay stations need a different voice again

In a discussion station, clear and steady is what you want. In a roleplay station with a distressed actor, the same delivery reads as cold. Warmth in the voice is part of what is being marked, and it lives in slowing down further, softening the volume slightly, and letting your pitch move rather than staying flat.

One practical warning: quieter in a roleplay does not mean inaudible. There is an assessor in the corner marking you who is further away than the actor, and if they cannot hear your empathy, it did not happen.

You cannot fix this without hearing yourself

None of the above is knowledge you lack. It is a habit problem, and habits only change when you get feedback. Record a two minute answer on your phone, from across the room rather than held at your face, and listen back for four things: did the volume hold to the end of each sentence, did you speed up as you went, did statements sound like statements, and could you hear thinking or only filling.

Then do it under pressure, because a calm recording tells you very little about what nerves do to your voice. MasterMed's live AI interviewer runs timed MMI stations and marks you against a rubric, which puts the clock and the stranger back into the equation so you can hear how you actually sound when it counts. The first speaking station is free on the trial, no card, and the trial never converts by itself.

Fix volume first, then sentence endings, then pace. In that order they take about a fortnight of attention, and they will lift the score of every answer you already know how to build.