What listeners extract from your voice
Listeners draw social conclusions from voice alone. Puts, Gaulin and Verdolini (2006) found that a lower-pitched, more masculine voice raised ratings of men's physical and social dominance. Ko, Sadler and Galinsky (2015) went further: when people were randomly assigned to a high- or low-rank role, their voices changed, and listeners who heard only the recordings could tell who held the higher rank.
The practical implication: the voice is not a neutral carrier of content. How you sound shapes how credible and confident you seem, alongside what you say.
Speaking rate: the WPM research
A common piece of advice is to slow down. The research is more mixed than that. In a classic experiment, Miller, Maruyama, Beaber and Valone (1976) played the same persuasive message at different speeds. Delivered fast, at about 195 words per minute, it was more persuasive and the speaker was seen as more credible than when delivered slowly, at about 100 words per minute. Very slow speech tends to come across as hesitant.
So the goal is not a slow voice but a controlled one: a pace that stays clear, with deliberate pauses at the ends of ideas rather than hurried filler. A speaker who can let a short pause sit, without rushing to fill it, sounds more in command than one who cannot.
Filler words, "um," "uh," "like," "you know", usually fill the moment while you plan what to say next. Some are normal in conversation and help listeners follow you; a steady stream of them makes a speaker sound less prepared. Reducing them is less about vocabulary than about being willing to pause silently instead.
Pitch and tonal variation
An animated voice is one of the three nonverbal tactics in Antonakis's charisma training (see what charisma actually is), and flat, monotone delivery tends to come across as disengaged. Variation in pitch and loudness marks emphasis, transitions and key points. More variation is not always better, though: in the Ko study above, speakers in high-rank roles actually varied their pitch less, but their loudness more.
Two specific patterns warrant attention. The first is upspeak, the rising intonation at the end of declarative sentences that turns statements into apparent questions. The linguist Robin Lakoff (1975) described this pattern as a way of hedging, and it is widely heard as uncertainty in formal and professional contexts, although linguists note it can also be used simply to check that the listener is following. The second pattern is the opposite problem: a flat, downward-resolved pitch at the end of every sentence, the vocal equivalent of a full stop that signals no further engagement. The most effective vocal delivery uses a mix of falling and level cadences, reserving the strong downward resolution for moments of genuine conclusion.
Resonance and projection
Pitch matters for perceived authority. Lower-pitched voices are rated as more dominant (Puts et al., 2006), and in hypothetical elections both men and women preferred female candidates with lower-pitched voices, a preference that held even for leadership roles usually held by women (Anderson & Klofstad, 2012). Note that this is about natural pitch differences between people; it does not mean you should force your voice lower.
What you can work on is the fullness of your normal voice. Breath support, posture and tension in the throat and jaw all affect it, and stress tends to raise pitch and tighten the voice. Relaxing that tension lets you speak in your natural range rather than a strained, higher one.
Projection, the capacity to fill a space without shouting, is similarly a function of breath support rather than volume. A well-projected voice carries on directed breath rather than on increased loudness. The distinction is perceptible to listeners: projection sounds calm and in control; shouting sounds effortful and dysregulated.
Puts, D.A. et al. (2006), "Dominance and the evolution of sexual dimorphism in human voice pitch." Evolution and Human Behavior, 27(4), 283–296. A masculine, lower-pitched voice raised ratings of men's physical and social dominance, the former more than the latter.
Ko, S.J. et al. (2015), "The sound of power: Conveying and detecting hierarchical rank through voice." Psychological Science, 26(1), 3–14. Speakers assigned to a high-rank role spoke with higher pitch, more variable loudness and less variable pitch; listeners used higher pitch and greater, more variable loudness to correctly infer rank.
Anderson, R.C. & Klofstad, C.A. (2012), "Preference for leaders with masculine voices holds in the case of feminine leadership roles." PLOS ONE, 7(12), e51216. In hypothetical elections, men and women preferred female candidates with lower-pitched voices, even for leadership roles typically held by women.
The state problem: why vocal exercises alone fail
Most vocal training approaches focus on the mechanics: breath exercises, resonance drills, rate control, filler-word monitoring. These have genuine value, but they share a fundamental limitation, they address the output without addressing the input.
The voice also reflects arousal. Under stress, pitch tends to rise, breathing becomes shallower and the voice tightens. These changes happen without deciding to make them, which is why technique practised in calm conditions is not always available when you are nervous.
This means that vocal training which does not address physiological state will produce the following pattern: the skill is present in low-pressure rehearsal and absent under social pressure, precisely the conditions where it matters most. The voice that sounds confident during practice collapses during the actual presentation because the internal state during the presentation is different from the internal state during practice.
Effective vocal development therefore requires two tracks: the technical track (mechanics, monitoring, specific exercises) and the state track (establishing a physiological baseline that the technical skills can actually operate from under pressure). The second track is the constraining variable. Internal coherence, the alignment of body, breath, and attention, is the prerequisite for sustained vocal presence, not a supplement to it.
Measurable targets
Vocal development responds well to objective measurement because the relevant variables are acoustic and therefore quantifiable. Specific metrics worth tracking include:
- Words per minute, tracked over time so you can see whether nerves push you much faster or slower than your normal pace
- Filler word frequency, as a percentage of total words, compared with your own earlier recordings
- Pitch range, standard deviation of F0 across a speech sample; flat delivery shows low SD; animated, engaging delivery shows higher SD
- Pause frequency and duration, number and length of deliberate pauses; effective speakers use pauses deliberately rather than filling all silence
Real-time vocal analysis
Charisma Coach AI analyses WPM, filler word frequency, fluency score, and pitch characteristics in real time, providing objective feedback on the specific vocal dimensions that matter. Compare it with Orai and Yoodli.
Scientific references
- Puts, D.A., Gaulin, S.J.C., & Verdolini, K. (2006). Dominance and the evolution of sexual dimorphism in human voice pitch. Evolution and Human Behavior, 27(4), 283–296.
- Ko, S.J., Sadler, M.S., & Galinsky, A.D. (2015). The sound of power: Conveying and detecting hierarchical rank through voice. Psychological Science, 26(1), 3–14.
- Anderson, R.C. & Klofstad, C.A. (2012). Preference for leaders with masculine voices holds in the case of feminine leadership roles. PLOS ONE, 7(12), e51216.
- Lakoff, R. (1975). Language and Woman's Place. Harper & Row.
- Miller, N., Maruyama, G., Beaber, R.J., & Valone, K. (1976). Speed of speech and persuasion. Journal of Personality and Social Psychology, 34(4), 615–624.